Intelligent puncture path planning method under ultrasonic guidance and system thereof

Through multi-scale deep learning and deep reinforcement learning technology, combined with dynamic risk assessment, the precise positioning and safety of ultrasonic guided puncture technology in complex anatomical structures is solved, and efficient and safe puncture path planning and real-time operation guidance are achieved to meet the needs of the The need for complex surgical environments.

CN119055330AActive Publication Date: 2024-12-03JILIN UNIV FIRST HOSPITAL

Patent Information

Application Number
CN202411186892.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-12-03
Estimated Expiration
2044-08-28

AI Technical Summary

Technical Problem

The existing ultrasound-guided puncture technology is difficult to achieve precise positioning and safe operation in complex anatomical structures. It lacks a real-time and dynamic risk assessment mechanism, and the real-time and robustness of the system are difficult to meet the needs of complex surgical environments.

Method used

Using multi-scale deep learning image segmentation and deep reinforcement learning for path planning, combined with dynamic risk assessment, augmented reality guidance interface is generated, and through real-time ultrasonic image processing, deep learning and reinforcement learning technology, doctors can provide accurate puncture paths and real-time operation guidance. .

Benefits of technology

The identification accuracy of complex anatomical structures is significantly improved, the accuracy and safety of path planning is achieved, and the introduction of a dynamic risk assessment mechanism has improved the safety of surgery, met the real-time needs of complex surgical environments, and reduced the difficulty and risks of operations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119055330A_ABST
    Figure CN119055330A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of intelligent puncture path planning methods, in particular to an intelligent puncture path planning method and system under ultrasonic guidance, and the method comprises the following steps: obtaining a real-time ultrasonic image sequence; preprocessing and enhancing the real-time ultrasonic image sequence; based on the enhanced ultrasonic image, performing multi-scale deep learning image segmentation to obtain a segmented anatomical structure label graph; target positioning and feature extraction are carried out based on the segmented anatomical structure label graph; based on the segmented anatomical structure label graph and the target position coordinates, path planning is conducted through deep reinforcement learning, and an optimal puncture path is obtained; performing dynamic risk assessment based on the optimal puncture path and the enhanced ultrasonic image to obtain a risk assessment result; and on the basis of the optimal puncture path, the risk assessment result and the enhanced ultrasonic image, an augmented reality guidance interface is generated, the optimal balance can be found among multiple targets, and the planning accuracy and safety are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent puncture path planning methods, and in particular to an intelligent puncture path planning method under ultrasound guidance and a system thereof. Background Art

[0002] In recent years, with the rapid development of minimally invasive surgical techniques, ultrasound-guided puncture surgery has played an increasingly important role in clinical practice. This technology brings less trauma, faster recovery time and lower risk of complications to patients. However, although ultrasound-guided technology has made significant progress, there are still many challenges in performing precise puncture operations in complex anatomical structures.

[0003] Traditional ultrasound-guided puncture methods mainly rely on the doctor's experience and skills. As Kaiyu Wu et al. pointed out in "Review of research on path planning and control methods of flexible steerable needle puncture robot" (Computer Assisted Surgery, 2022, 27(1):91-112), conventional rigid straight needles are difficult to bypass obstacles during puncture and cannot adapt to complex anatomical structures. This method is often difficult to achieve precise positioning and safe operation when facing deep organs or needing to avoid important blood vessels and nerves.

[0004] To solve this problem, researchers have proposed various intelligent path planning methods. Li et al. introduced an ultrasound-guided puncture robot based on adaptive control and reinforcement learning in "An Adaptive Control Method and Learning Strategy for Ultrasound-Guided Puncture Robot" (Electronics, 2024, 13:580). Although this method has improved the accuracy of puncture to a certain extent, it still has some limitations. First, this method mainly focuses on the control strategy of the robot, but pays insufficient attention to complex anatomical structure recognition and risk assessment. Secondly, the real-time performance of this method needs to be improved, and it is difficult to meet the needs of dynamic surgical environments.

[0005] In addition, although some existing deep learning-based methods have made progress in image recognition, they often ignore the dynamic risk assessment during surgery. These methods usually use static pre-training models, which are difficult to adapt to the challenges brought by tissue deformation and individual differences in patients during surgery.

[0006] In summary, the existing technology still has obvious deficiencies in the following aspects:

[0007] 1. Lack of accurate recognition and segmentation capabilities for complex anatomical structures;

[0008] 2. It is difficult for the path planning algorithm to strike a good balance between safety, efficiency and operability;

[0009] 3. Lack of real-time and dynamic risk assessment mechanism;

[0010] 4. The real-time and robustness of the system cannot meet the requirements of complex surgical environments.

[0011] In response to these problems, the present invention proposes an intelligent puncture path planning method and system under ultrasound guidance. Summary of the invention

[0012] The purpose of the present invention is to provide an intelligent puncture path planning method and system under ultrasound guidance to solve the above problems.

[0013] The present invention provides an intelligent puncture path planning method under ultrasound guidance, comprising the following steps: obtaining a real-time ultrasound image sequence; preprocessing and enhancing the real-time ultrasound image sequence to obtain an enhanced ultrasound image; performing multi-scale deep learning image segmentation based on the enhanced ultrasound image to obtain a segmented anatomical structure label map; performing target positioning and feature extraction based on the segmented anatomical structure label map to obtain target position coordinates and feature descriptors; performing path planning using deep reinforcement learning based on the segmented anatomical structure label map and the target position coordinates to obtain an optimal puncture path; performing dynamic risk assessment based on the optimal puncture path and the enhanced ultrasound image to obtain a risk assessment result; and generating an augmented reality guidance interface based on the optimal puncture path, the risk assessment result and the enhanced ultrasound image.

[0014] Specifically, the step of preprocessing and enhancing the real-time ultrasonic image sequence includes: calculating the local standard deviation and the global standard deviation according to the local area of ​​each pixel point in the real-time ultrasonic image sequence; determining the adaptive filtering window size based on the ratio of the local standard deviation to the global standard deviation; using the adaptive filtering window size, performing median filtering on the real-time ultrasonic image sequence to obtain a denoised image sequence; calculating a local histogram for each image in the denoised image sequence; determining an adaptive clipping threshold according to the maximum and minimum values ​​of the local histogram; using the adaptive clipping threshold, performing contrast-limited adaptive histogram equalization on the denoised image sequence to obtain an equalized image sequence; calculating the local mean and the global mean of each image in the equalized image sequence; determining an adaptive Gamma correction parameter based on the difference between the local mean and the global mean; using the adaptive Gamma correction parameter, performing Gamma correction on the equalized image sequence to obtain the enhanced ultrasonic image.

[0015] Specifically, the step of performing multi-scale deep learning image segmentation based on the enhanced ultrasound image includes: constructing a multi-scale U-Net++ network model, including an encoder and a decoder; in the encoder, extracting the multi-scale features of the enhanced ultrasound image through continuous convolution layers and pooling layers; in the decoder, gradually restoring the spatial resolution of the feature map through transposed convolution layers; adding dense jump connections between the encoder and the decoder to achieve the fusion of features of different scales; calculating the spatial attention map of the fused feature map, which reflects the importance of different areas of the image; calculating the channel attention vector of the fused feature map, which reflects the importance of different feature channels; multiplying the spatial attention map and the channel attention vector with the fused feature map to obtain a weighted feature map; based on the weighted feature map, generating a multi-category segmentation probability map through a 1x1 convolution layer and a softmax activation function; calculating Dice loss, boundary loss and Focal loss, and weightedly combining these three losses to obtain a total loss function; optimizing the parameters of the network model through back propagation and gradient descent, and finally obtaining the segmented anatomical structure label map.

[0016] Specifically, the step of performing target positioning and feature extraction based on the segmented anatomical structure label map includes: obtaining the segmented anatomical structure label maps of the current frame and the previous k frames; performing temporal consistency analysis on the k+1 frame label map to screen out stable connected areas; calculating the area, perimeter and shape complexity of the stable connected areas, and screening out candidate target areas according to a preset threshold; calculating weights for pixels in the candidate target area according to their grayscale values ​​in the enhanced ultrasound image; using the weights, calculating the weighted center of mass of the candidate target area to obtain the target position coordinates; extracting a local area of ​​a fixed size with the target position coordinates as the center; calculating shape features for the local area, including roundness, eccentricity and directionality; calculating texture features for the local area, including grayscale co-occurrence matrix features and local binary pattern features; using a pre-trained convolutional neural network to extract deep features of the local area as context features; and splicing the shape features, the texture features and the context features to form the feature descriptor.

[0017] Specifically, the step of using deep reinforcement learning for path planning based on the segmented anatomical structure label map and the target position coordinates includes: calculating the distance from each position to the nearest obstacle according to the segmented anatomical structure label map to obtain an obstacle distance map; generating an uncertainty map based on the confidence of the segmented anatomical structure label map; estimating the tissue deformation field using the registration results between consecutive frames to obtain a deformation degree map; calculating a local vascular density map according to the vascular segmentation result; weightedly combining the obstacle distance map, the uncertainty map, the deformation degree map and the vascular density map to obtain a comprehensive risk map; constructing a reinforcement learning environment based on the comprehensive risk map and the target position coordinates; in the reinforcement learning environment, using an improved Deep Q-Network algorithm for global path planning; based on the global path, using a newly designed Advantage The Actor-Critic algorithm is used to optimize the local path; an adaptive reward function is designed according to the safety, efficiency and accuracy of the path; multiple agents are introduced, each of which is responsible for different subtasks, such as safety, efficiency and accuracy; a knowledge distillation mechanism is used to enable knowledge sharing and collaborative learning among the agents; an integrated teacher strategy is constructed based on the strategies of all agents; the integrated teacher strategy is used to guide the learning process of each agent; combined with the learned Q value, an improved Rapidly-exploring Random Trees algorithm is used for path sampling and optimization; the optimization is iterated until convergence or a preset number of iterations is reached to obtain the optimal puncture path.

[0018] Specifically, the steps of constructing the dynamic risk map include: initializing weight coefficients based on the obstacle distance map, the uncertainty map, the deformation degree map and the blood vessel density map; calculating the difference between the current risk map and the target risk distribution; updating the weight coefficients based on the difference using the gradient descent method; recalculating the comprehensive risk value using the updated weight coefficients; repeating the above steps until the risk map converges or reaches a preset number of iterations; and outputting the final dynamic risk map.

[0019] Specifically, the step of performing dynamic risk assessment based on the optimal puncture path and the enhanced ultrasound image includes: calculating the path curvature along the optimal puncture path; determining the sampling interval based on the path curvature and the dynamic risk map; extracting path key points according to the sampling interval; for each path key point, extracting obstacle distance, uncertainty, tissue deformation and vascular density information in the local area; using fuzzy logic rules, converting the extracted information into a fuzzy set; based on evidence theory, calculating the basic probability distribution of each risk factor; using the Dempster combination rule, fusing the evidence of multiple risk factors to obtain a local risk assessment result; constructing a recursive neural network and taking the local risk assessment result sequence as input; extracting the global features of the optimal puncture path; splicing the global features with the hidden state of the recursive neural network; and outputting the overall path risk assessment result through a fully connected layer and a Sigmoid activation function.

[0020] Specifically, the step of generating an augmented reality guidance interface based on the optimal puncture path, risk assessment results and enhanced ultrasound images includes: calculating the saliency map and contrast map of the enhanced ultrasound image; generating an adaptive fusion weight map based on the saliency map and the contrast map; using the adaptive fusion weight map, superimposing the optimal puncture path on the enhanced ultrasound image; calculating the global mean and standard deviation of the dynamic risk map; determining an adaptive risk threshold based on the global mean and the standard deviation; marking high-risk areas on the superimposed image based on the adaptive risk threshold; constructing a Kalman filter to initialize the needle tip state estimation; extracting the needle tip feature template in the enhanced ultrasound image; using the normalized cross-correlation method to perform template matching in the predicted area; updating the needle tip position estimation by combining the Kalman filter prediction and template matching results; calculating the distance from the needle tip to the target, the current local risk value and the needle tip movement speed; dynamically adjusting the color and transparency of the visual feedback based on the above calculation results; and at the same time, adjusting the frequency and volume of the audio feedback based on the above calculation results; synthesizing all visual elements into a single layer to generate the final augmented reality guidance interface.

[0021] Specifically, before executing the method, the following data preparation steps are also included: obtaining the model and parameter information of the ultrasound equipment to ensure that its frame rate is not less than 30fps and the resolution is not less than 1024x768; collecting relevant medical information of the patient, including age, gender, medical history and relevant examination results; obtaining the surgical plan, including puncture target, entry point selection and avoidance structure; loading pre-trained deep learning model parameters, including image segmentation model and feature extraction model; initializing the policy network and value network parameters of the reinforcement learning agent; calibrating the spatial position and direction of the ultrasound probe; and performing system self-checks to ensure that all modules are working properly.

[0022] The ultrasonic-guided intelligent puncture path planning system for implementing the method comprises: an image acquisition module for obtaining a real-time ultrasonic image sequence; an image preprocessing module for preprocessing and enhancing the real-time ultrasonic image sequence; an image segmentation module for performing multi-scale deep learning image segmentation on the enhanced ultrasonic image; a target positioning module for performing target positioning and feature extraction based on the segmentation result; a path planning module for performing path planning using deep reinforcement learning; a risk assessment module for performing dynamic risk assessment; and a visual guidance module for generating an augmented reality guidance interface; wherein the image acquisition module comprises an ultrasonic probe and an image acquisition card; the image preprocessing module, the image segmentation module, the target positioning module, the path planning module and the risk assessment module are all implemented by a computer with a GPU acceleration function; and the visual guidance module comprises a display and a stereo sound system.

[0023] The method first obtains a real-time ultrasound image sequence, preprocesses and enhances it, and then uses multi-scale deep learning to perform image segmentation and accurately locate targets and key structures. Next, path planning is performed through a deep reinforcement learning algorithm, and the puncture strategy is adjusted in real time in combination with a dynamic risk assessment mechanism. Finally, the system generates an intuitive augmented reality guidance interface to provide doctors with real-time and accurate operation guidance.

[0024] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0025] 1. Through the multi-scale deep learning segmentation algorithm, the recognition accuracy of complex anatomical structures is significantly improved, laying the foundation for accurate path planning;

[0026] 2. Using deep reinforcement learning for path planning can find the optimal balance between multiple goals, improving the accuracy and safety of planning;

[0027] 3. The introduction of a dynamic risk assessment mechanism enables real-time assessment and adjustment of surgical risks, greatly improving the safety of surgery;

[0028] 4. Through optimized hardware configuration and parallel computing technology, the real-time operation of complex algorithms is realized, meeting the needs of dynamic surgical environments;

[0029] 5. Integrates augmented reality technology to provide doctors with intuitive and real-time operation guidance, effectively reducing the difficulty and risk of operation.

[0030] In general, the present invention provides a comprehensive and efficient solution for ultrasound-guided precision puncture surgery, which is expected to significantly improve the success rate and safety of minimally invasive surgery and bring better treatment effects to patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The figure is an overall logic block diagram of the method of the present invention.

[0032] Figure 2 It is a logic block diagram of the image preprocessing and enhancement steps of the present invention.

[0033] Figure 3 This is a logic block diagram of the multi-scale deep learning image segmentation steps of the present invention.

[0034] Figure 4 This is a logic block diagram of the deep reinforcement learning path planning steps of the present invention.

[0035] Figure 5 It is a logic block diagram of the dynamic risk assessment steps of the present invention.

[0036] Figure 6 This is a logic block diagram of the steps of generating an augmented reality guidance interface of the present invention.

[0037] Figure 7 It is the overall logic block diagram of the system of the present invention.

[0038] Figure 8 It is a multi-dimensional dynamic decision tree diagram of the present invention and comparative examples 1-3.

[0039] Fig. 9 It is a comparison diagram of the 3D puncture trajectory of the present invention and comparative examples 1-3. DETAILED DESCRIPTION

[0040] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0041] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0042] Please refer to Figure 1-9 The present invention discloses an intelligent puncture path planning method and system under ultrasound guidance, which is particularly suitable for precise puncture surgery of deep organs such as liver and prostate. The method provides doctors with safe and efficient puncture path planning and real-time guidance by combining real-time ultrasound image processing, deep learning and reinforcement learning technology. The specific implementation methods of the present invention will be described in detail below.

[0043] Step S1: Obtain a real-time ultrasound image sequence.

[0044] First, the present invention uses a high-frequency ultrasound probe to collect a real-time ultrasound image sequence of the target area. In a preferred embodiment, we use a Philips iU22 ultrasound system with a frequency range of 5-12MHz. In order to ensure image quality and real-time performance, we set the acquisition frame rate to no less than 30fps and the resolution to no less than 1024x768. The selection of these parameters is based on clinical practice experience, which can meet the needs of real-time processing while ensuring image details.

[0045] Step S2: Preprocessing and enhancing the real-time ultrasound image sequence

[0046] This step aims to improve the quality of ultrasound images and lay the foundation for subsequent analysis. First, we use the adaptive median filter algorithm to denoise the image. The core of this algorithm is to dynamically adjust the filter window size to adapt to the noise level in different areas. Specifically, we calculate the local standard deviation σ of each pixel local (x,y) and the global standard deviation σ global To determine the appropriate window size. The formula for calculating the window size is:

[0047]

[0048] Here, we limit the window size to between 3 and 11. This range is based on empirical values ​​derived from a large number of experiments and can achieve a good balance between denoising effect and detail preservation.

[0049] Next, we perform contrast-limited adaptive histogram equalization (CL-AHE) on the denoised image. The purpose of this step is to enhance the image contrast while avoiding over-enhancing the noise. To this end, we introduce the concept of adaptive clipping threshold, which is calculated as:

[0050] P clip(x,y)=α+(1-α)·(max(H(x,y))-min(H(x,y))) / max(H(x,y))

[0051] Among them, H(x, y) is the local histogram, α is the basic clipping threshold, and the default value is set to 0.01. This value is obtained through repeated experimental optimization and can achieve a good balance between enhancing image contrast and suppressing noise enhancement.

[0052] Finally, we perform adaptive gamma correction to further optimize the brightness and contrast of the image. The calculation formula for the gamma correction parameters is as follows

[0053] γ(x, y) = 1 + β·(μ local (x, y)-μ global ) / σ global

[0054] Here, μ local (x, y) is the local mean, μ global is the global mean, β is the adjustment parameter, and the default value is 0.5. The choice of this parameter value is based on the trade-off between contrast and brightness, and can be fine-tuned according to the specific application scenario.

[0055] Through the above steps, we obtain the enhanced ultrasound image I with significantly improved quality. enhanced (t), which lays a solid foundation for subsequent image segmentation and feature extraction.

[0056] Step S3: Perform multi-scale deep learning image segmentation based on the enhanced ultrasound image.

[0057] In this step, we use the improved U-Net++ network model for image segmentation. First, we build an encoder-decoder network structure with 5 scales. The encoder uses ResNet-50 with dilated convolution as the backbone network to increase the receptive field and capture more contextual information. The decoder uses transposed convolution for upsampling and fuses features of different scales through dense skip connections to retain more detailed information.

[0058] To further improve the performance of the model, we introduced spatial and channel attention mechanisms in the feature fusion module. Spatial attention is calculated through 3x3 convolution, which can highlight the key areas in the image; while channel attention is implemented using global average pooling and 1x1 convolution, which helps to emphasize important feature channels. The introduction of this dual attention mechanism significantly improves the model's ability to recognize key anatomical structures.

[0059] In the design of the loss function, we adopted a multi-task learning strategy, taking into account the overall segmentation accuracy, boundary accuracy and category balance. Specifically, our loss function consists of three parts:

[0060] L total =λ1·L dice +λ2·L boundary +λ3·L focal

[0061] Among them, L dice is the Dice loss, which is used to optimize the overall segmentation accuracy; L boundary is the boundary loss, focusing on improving the accuracy of boundary segmentation; L focal is the Focal loss, which is used to deal with the problem of class imbalance. λ1, λ2, and λ3 are weight coefficients, and the default values ​​are set to 0.5, 0.3, and 0.2. These weight values ​​are obtained through a large number of experimental optimizations and can achieve a good balance between different task objectives.

[0062] Through this multi-scale, multi-task deep learning method, we can obtain high-quality segmentation results L(t), which provides reliable anatomical structure information for subsequent target positioning and path planning.

[0063] Step S4: Based on the segmented anatomical structure label map, target positioning and feature extraction are performed.

[0064] In this step, we first need to accurately locate the target area. To this end, we designed an improved connected domain analysis algorithm, which not only considers the segmentation results of the current frame, but also integrates the information of the previous k frames (default k = 5) to improve the stability of positioning. Specifically, we first perform pixel-by-pixel AND operations on the k+1 frame label map to obtain a stable region mask. Then, we perform morphological opening operations on this mask to remove small noise areas. Next, we calculate the area, perimeter, and shape complexity of the remaining connected regions. Finally, the candidate target areas are screened out according to the preset thresholds (for example, the area is greater than 100 pixels and the shape complexity is less than 0.8).

[0065] After determining the candidate target area, we use the weighted centroid algorithm to accurately locate the target position. Here, we calculate the weight according to the gray value of the pixels in the candidate area in the enhanced ultrasound image:

[0066]

[0067] We then use these weights to calculate the weighted centroid and get the target position coordinates P target :

[0068]

[0069] This weighted centroid method takes into account the grayscale distribution of pixels in the target area and can locate the center of the target more accurately, especially for targets with irregular shapes or complex internal structures.

[0070] Next, we perform feature extraction. Centered on the target location coordinates, we extract a local region of 64x64 pixels. This size is chosen to strike a balance between including sufficient context information and maintaining computational efficiency. Within this local region, we extract three types of features:

[0071] The first is shape features, including roundness, eccentricity, and directionality, etc. These features can describe the basic geometric properties of the target and are very helpful for distinguishing different types of anatomical structures (such as blood vessels, tumors, etc.).

[0072] The second is texture features. We use the gray-level co-occurrence matrix (GLCM) to calculate statistics such as energy, contrast, and correlation. The calculation of GLCM usually considers 4 directions (0°, 45°, 90°, 135°) and a distance of 1-5 pixels to capture texture information of different scales and directions. In addition, we also use local binary patterns (LBP) to extract texture features. LBP features are rotationally invariant and can effectively describe local texture patterns.

[0073] Finally, we extract deep features. Here we use a pre-trained ResNet-18 network to extract 512-dimensional deep features. ResNet-18 is chosen because it has relatively low computational complexity while maintaining high feature extraction capabilities, which is suitable for real-time processing needs. The pre-trained model uses weights trained on large-scale natural image datasets, and we found that these general features also perform well in medical image analysis tasks.

[0074] By stitching these features together, we get a comprehensive feature descriptor D target This feature descriptor contains the shape, texture, and semantic information of the target, providing rich feature representation for subsequent path planning.

[0075] Step S5: Based on the segmented anatomical structure label map and target position coordinates, deep reinforcement learning is used for path planning.

[0076] This step is the core innovation of the present invention. We use an improved deep reinforcement learning algorithm for path planning. First, we need to build a dynamic risk map. The formula for building the risk map is as follows:

[0077] R(xy,t)=w1(t)·D(x,y,t)+w2(t)·U(x,y,t)+w3(t)·C(x,y,t)+w4(t)·V(x,y,t)

[0078] Where R(x, y, t) is the comprehensive risk value at position (x, y) and time t, D(x, y, t) represents the distance to the nearest obstacle, U(x, y, t) represents uncertainty, C(x, y, t) represents the degree of tissue deformation, and V(x, y, t) represents the vascular density. w1(t), w2(t), w3(t), and w4(t) are dynamically adjusted weight coefficients. In order to dynamically adjust these weights, we use a gradient descent-based method:

[0079]

[0080] Here, w i (t) is the i-th weight at time t, is the gradient of the risk value R to the weight, η is the learning rate, usually set between 0.01 and 0.1, R target is the target risk value, R current is the current risk value. This dynamic adjustment mechanism enables the system to continuously optimize risk assessment strategies based on real-time feedback.

[0081] Next, we adopt a hierarchical reinforcement learning framework for path planning. At the global planning level, we use the improved Deep Q-Network (DQN) algorithm

[0082] Q(sa)=(1-α)Q(s,a)+α[r+γ·max(Q′(s′,a′)+ε·H(π(s′)))]

[0083] Here, Q(s, a) is the Q-value of taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, Q′(s′, a′) is the Q-value of taking action a′ in the next state s′, ε is the entropy regularization coefficient, and H(π(s′)) is the entropy of strategy π in state s′. Introducing strategy entropy can encourage exploration and prevent the algorithm from converging to a suboptimal solution too early.

[0084] At the local optimization level, we designed a new Advantage Actor-Critic (A2C) algorithm:

[0085] A(sa)=Q(s,a)-V(s)

[0086]

[0087]

[0088] Among them, A(s, a) is the advantage function, V(s) is the state value function, θ π are policy network parameters, θ v is the value network parameter, J(θ) is the policy objective function, L(θ) is the value function loss, and b(s) is the baseline function used to reduce variance. This actor-critic structure can improve learning efficiency while ensuring stability.

[0089] In order to further improve the performance of the algorithm, we introduced a multi-agent collaborative learning mechanism. Each agent is responsible for different subtasks (such as security, efficiency, etc.), and collaborative learning is carried out through the knowledge distillation mechanism:

[0090] In order to further improve the performance of the algorithm, we introduced a multi-agent collaborative learning mechanism. Each agent is responsible for different subtasks (such as security, efficiency, etc.), and collaborative learning is carried out through the knowledge distillation mechanism:

[0091] Q i (sa) = Q i (s,a)+α i [r i +γ·max(Q i (s′, a′))-Q i (s,a)]+β i ·KL(π teacher ||π i )

[0092] Here, Q i (s, a) is the Q value of the ith agent taking action a in state s, α i is the learning rate of the ith agent, r i is the reward obtained by the ith agent, β i is the knowledge distillation coefficient, KL(π teacher ||π i ) is the KL divergence between the teacher’s strategy and the strategy of agent i. The teacher’s strategy is obtained by integrating the strategies of all agents:

[0093] π teacher (a|s)=softmax(∑w i Q i (s, a) / τ)

[0094] Among them, π teacher (a|s) is the probability that the teacher strategy chooses action a in state s, and w i The weight of the i-th agent, τ is the temperature parameter, which controls the degree of “softness” of the strategy.

[0095] Finally, we use the improved Rapidly-exploring RandomTrees (RRT*) algorithm for path optimization. During the sampling process, we combine the learned Q value for biased sampling:

[0096] P(x random )∝exp(Q(s x , a optimal ) / T)

[0097] Here, P(x random ) is to select a random point X random The probability of Q(s x , a optimal ) is the state s x The Q value of the optimal action under the condition of , T is the temperature parameter, which controls the balance between exploration and exploitation. In this way, we can use the knowledge gained from reinforcement learning to guide sampling and improve planning efficiency while maintaining the rapid exploration capability of the RRT* algorithm.

[0098] Through the above steps, we finally get the optimal puncture path Path optimal This path not only takes into account spatial geometric constraints, but also integrates multiple factors such as safety and efficiency, and can provide doctors with the best puncture recommendations.

[0099] Step S6: Perform dynamic risk assessment based on the optimal puncture path and the enhanced ultrasound image.

[0100] After obtaining the optimal puncture path, we need to perform a dynamic risk assessment on it to ensure the safety of the surgical process. First, we use an adaptive sampling algorithm to extract key points of the path. The core idea of ​​this algorithm is to dynamically determine the sampling interval based on the path curvature and risk distribution. Specifically, in areas with large path curvature or high risk, we will increase the sampling density; conversely, in straight segments or low-risk areas, we will appropriately reduce the sampling points to balance computational efficiency and evaluation accuracy.

[0101] For each key point sampled, we perform a local risk assessment. This assessment takes into account multiple factors, including obstacle distance, uncertainty, tissue deformation, and vascular density. We use a combination of fuzzy logic and evidence theory to deal with the uncertainty of these factors. First, we convert each risk factor into a fuzzy set using fuzzy rules. For example, for obstacle distance, we may have fuzzy sets such as "very close", "close", "medium", "far", and "very far". Then, we calculate the basic probability assignment (BPA) of each risk factor based on evidence theory. Finally, the evidence of multiple risk factors is fused through the Dempster combination rule to obtain a comprehensive local risk assessment result.

[0102] The advantage of this approach is that it can effectively handle the uncertainty and ambiguity commonly found in medical images, providing more robust and interpretable risk assessment results.

[0103] In order to obtain the risk assessment of the overall path, we designed a risk aggregation function based on a recurrent neural network (RNN). Specifically, we use a long short-term memory network (LSTM) as the basic unit of the RNN and take the sequence of local risk assessment results as input:

[0104] h t =LSTM(Risk local (P t ), h t-l )

[0105] Among them, h t is the hidden state at time t, Risk local (P t ) is the local risk assessment result of the key point at time t.

[0106] In addition to the local risk sequence, we also extract the global features of the path, such as path length, average curvature, etc. These global features are spliced ​​together with the final hidden state of LSTM, and the overall path risk assessment result is output through the fully connected layer and Sigmoid activation function:

[0107] Risk global =σ(W·[h final , Path features ]+b)

[0108] Among them, Risk global is the global risk assessment result, Path features is the path global feature, σ is the Sigmoid function, W and b are the weight matrix and bias term respectively, h final is the final hidden state of the LSTM.

[0109] This approach combining local and global information enables a comprehensive assessment of path risk, taking into account the local risk of each point on the path while not neglecting the overall characteristics of the path.

[0110] Step S7: Generate an augmented reality guidance interface based on the optimal puncture path, risk assessment results and enhanced ultrasound images.

[0111] The last step is to integrate all the information and generate an intuitive augmented reality guidance interface to assist doctors in performing precise punctures. First, we need to superimpose the planned optimal puncture path onto the real-time ultrasound image. To this end, we designed an improved image fusion algorithm that takes into account image features and human eye perception characteristics:

[0112] I fusion =λ(x, y)·I enhanced +(1-λ(x,y))·I path

[0113] Among them, I fusion is the fused image, I enhanced is the enhanced ultrasound image, I path is the path image, λ(x, y) is the adaptive fusion weight, calculated by the following formula:

[0114] λ(xy)=sigmoid(α·S(x,y)+β·C(x,y))

[0115] Here, S(x, y) is the saliency map, C(x, y) is the contrast map, and α and β are weight coefficients. This adaptive fusion method can clearly display the planned path while maintaining the details of the original ultrasound image.

[0116] Next, we need to mark the high-risk areas. We use an adaptive threshold algorithm to dynamically determine the threshold of high-risk areas based on the overall risk distribution:

[0117] Threshold risk =μ risk +k·σ risk

[0118] Among them, μ risk and σ risk are the mean and standard deviation of the risk value, respectively, and k is an adjustable parameter, usually set between 1.5 and 2.5. Areas above this threshold will be marked as high-risk areas and marked with a striking color (such as red) on the interface.

[0119] In order to achieve real-time needle tip tracking, we designed a robust needle tip tracking algorithm that combines Kalman filtering and template matching. Kalman filtering is used to predict the motion state of the needle tip:

[0120] X t =F·X t-1 +B·u t +w t

[0121] Z t =H·X t +v t

[0122] Among them, X t It is the needle tip state, Z t is the observation value, F is the state transfer matrix, H is the observation matrix, w t and v tare process noise and observation noise respectively, B is the control input matrix, u t is the control input.

[0123] Based on the prediction, we use the template matching method for precise positioning. The matching score is calculated by the normalized cross-correlation coefficient (NCC):

[0124] Among them, Template is a predefined needle tip template, I roi (x, y) is the image patch of the region of interest. Finally, we designed a multimodal feedback system to dynamically adjust the visual and audio feedback strength according to the puncture status:

[0125] Score(x,y)=NCC(Template,I roi (x, y)

[0126] Visual feedback =h v (dist to_target ,Risk local , needle orientation )

[0127] Audio feedback =h a (dist to_target ,Risk local , needle speed )

[0128] Where Score(x, y) is the matching score of position (x, y), NCC is the normalized cross-correlation coefficient, Template is the predefined tip template, and I roi (x, y) is the image of the region of interest at position (x, y). Visual_feedback is the visual feedback strength, ranging from [0, 1], where 0 means no visual feedback and 1 means the strongest visual feedback; Audio_feedback is the audio feedback strength, ranging from [0, 1], where 0 means no audio feedback and 1 means the maximum volume audio feedback; h v is the nonlinear mapping function of visual feedback, which maps the input parameters to the visual feedback intensity. a is a nonlinear mapping function for audio feedback, mapping the input parameters to the audio feedback intensity. For example, when approaching a target or entering a high-risk area, the visual cue becomes more obvious and the audio warning is enhanced accordingly. to_target is the Euclidean distance from the needle tip to the target position, in millimeters (mm); Risk local The local risk value of the current position, ranging from [0, 1], 0 represents the lowest risk, and 1 represents the highest risk. needle orientationis the direction vector of the needle, expressed as a three-dimensional unit vector [x, y, z]. speed is the moving speed of the needle, in millimeters per second (mm / s).

[0129] These nonlinear mapping functions h v and h a Can be defined as:

[0130] h v (dist,risk,orient)=sigmoid(w1·(1-dist / d max )+w2·risk+w3·(1-dot(orient,target o rient)))

[0131] h a (dist, risk, speed) = sigmoid (w4·(1-dist / d max )+w5·risk+w6·(speed / speed max ))

[0132] Among them, sigmoid(x) = 1 / (1+exp(-x)) is the Sigmoid function, which is used to map the output to the range of [0, 1]. w1, w2, w3, w4, w5, w6 are weight parameters, which are used to adjust the importance of each factor. max is the preset maximum distance threshold. When dist_to_target exceeds this value, the distance factor is 0. det(orient, target_orient) is the dot product of the current direction of the needle and the target direction, which is used to measure the degree of direction matching. target_orient is the target direction, expressed as a three-dimensional unit vector [x, y, z]. speed max is the preset maximum speed threshold. When needle_speed exceeds this value, the speed factor is 1. These definitions enable the feedback system to dynamically adjust the intensity of visual and audio feedback based on the distance from the needle tip to the target, local risk, direction of the needle, and speed. For example, when the needle tip approaches the target (dist_to_target decreases), enters a high-risk area (Risk_local increases), the direction of the needle deviates from the target (det(orient, target_orient) decreases), or the speed is too fast (needle_speed increases), the system will increase the corresponding feedback intensity. In this way, the system can provide the operator with intuitive and real-time multimodal feedback, effectively improving the accuracy and safety of the surgery.

[0133] Through the above steps, we generated a comprehensive augmented reality guidance interface, which not only displays the optimal puncture path and real-time needle tip position, but also provides intuitive risk warnings through multimodal feedback. This interface can greatly improve the doctor's operating accuracy and confidence, thereby significantly improving the surgical outcome.

[0134] In summary, the present invention provides a comprehensive and innovative ultrasound-guided intelligent puncture path planning method and system by combining advanced image processing technology, deep learning, reinforcement learning, and augmented reality technology. This method can not only accurately locate the target and plan the optimal path, but also assess the risk in real time and provide intuitive operation guidance. It is expected to play an important role in clinical practice and improve the safety and success rate of minimally invasive surgery.

[0135] After the detailed description of the above method steps, we now specifically introduce the system for implementing this method. The ultrasound-guided intelligent puncture path planning system of the present invention is a complex system integrating multiple functional modules, each of which plays a key role in the entire workflow.

[0136] First, the core of the system is the image acquisition module 1. This module consists of a high-frequency ultrasound probe and a high-performance image acquisition card. In one of our embodiments, we chose the Philips iU22 ultrasound system equipped with a 5-12MHz linear array probe. This probe can provide high-resolution real-time images, which is particularly suitable for fine scanning of deep organs. It is matched with an image acquisition card that supports high frame rate acquisition, such as the Matrox Radient eV-CXP frame grabber, which can achieve an image acquisition rate of up to 60fps, fully meeting our real-time processing needs.

[0137] Next is the image preprocessing module 2. This module is responsible for denoising and enhancing the original ultrasound image. Considering the particularity and complexity of medical image processing, we use GPU acceleration technology in this module. Specifically, we use NVIDIA Tesla V100 GPU to execute image processing algorithms. This high-performance GPU has 5120 CUDA cores and 640 Tensor cores, which can greatly accelerate the running speed of our algorithms such as adaptive median filtering and contrast-limited adaptive histogram equalization, ensuring the real-time performance of the system.

[0138] The image segmentation module 3 is another key component of the system. This module runs the multi-scale U-Net++ network that we described in detail earlier. In order to support the efficient operation of deep learning models, we also use GPU acceleration. In practice, we found that a single GPU may not be able to meet the needs of real-time segmentation, so we adopted a multi-GPU parallel computing solution. Specifically, we used two NVIDIA RTX 3090 GPUs, each equipped with 24GB of video memory, which is enough to accommodate our large neural network model. Through PyTorch's DistributedDataParallel function, we achieved efficient parallel training and inference of the model.

[0139] The target positioning module 4 and the path planning module 5 are both computationally intensive components. These two modules mainly run on the CPU because they involve more logical judgments and complex mathematical operations. In our system, we chose the Intel Xeon Gold 6248R processor, which is a server-level CPU with 24 cores and 48 threads and a main frequency of up to 4.0GHz. This high-performance CPU can quickly process complex connected domain analysis, feature extraction, and reinforcement learning algorithms.

[0140] The risk assessment module 6 is a complex module that integrates multiple technologies. It needs to process image data and numerical calculations at the same time. Therefore, we adopted the collaborative work of CPU and GPU. Image-related processing tasks, such as the generation of local risk maps, are performed on the GPU; while risk aggregation based on recurrent neural networks is performed on the CPU. This heterogeneous computing architecture fully utilizes the advantages of different hardware and improves the overall computing efficiency.

[0141] Finally, the visual guidance module 7 is responsible for generating the augmented reality interface. The output of this module is directly facing the user, so it has high requirements for real-time and visual quality. We use OpenGL for efficient graphics rendering, and through NVIDIA's CUDA-OpenGL interoperability function, we achieve fast transmission of GPU calculation results to graphics display. For output devices, we chose a 4K monitor with a resolution of 3840x2160 and a refresh rate of 144Hz to provide a clear and smooth visual experience. At the same time, we also equipped a high-quality stereo sound system to provide accurate audio feedback.

[0142] It is worth mentioning that the modules of the whole system transmit data through a high-speed internal network. We use InfiniBand network technology, which has a bandwidth of up to 200Gb / s, to ensure the real-time transmission of large amounts of image and computing data and minimize system latency.

[0143] In addition, considering the particularity of the medical environment, our system also has a complete fault tolerance and backup mechanism. All key hardware is equipped with redundant design and uses a RAID storage system to protect important data. The system is also equipped with an uninterruptible power supply (UPS) to cope with possible power failures.

[0144] In general, this ultrasound-guided intelligent puncture path planning system is a highly integrated and powerful medical assistance system. It perfectly combines advanced hardware technology with innovative software algorithms to provide doctors with a reliable, efficient and intuitive surgical assistance tool. Through this system, we hope to significantly improve the accuracy and safety of minimally invasive surgery and ultimately benefit more patients.

[0145] In order to verify the superiority of the ultrasound-guided intelligent puncture path planning method and system of the present invention, we designed a series of comparative experiments. These experiments not only demonstrate the effect of our method, but also make a comprehensive comparison with the prior art. Below we will introduce the settings of the embodiment and comparative example in detail, as well as the corresponding experimental results and analysis.

[0146] Embodiment 1: The method of the present invention

[0147] In this embodiment, we fully implemented the method proposed in the present invention, including all steps such as adaptive image enhancement, multi-scale deep learning segmentation, path planning based on deep reinforcement learning, and dynamic risk assessment. We used the hardware configuration described above, including Philips iU22 ultrasound system, NVIDIA Tesla V100 GPU, etc.

[0148] Comparative Example 1: Traditional Manual Planning Method

[0149] This comparison represents the traditional method currently widely used in clinical practice. In this method, doctors rely entirely on their own experience and judgment to plan the puncture path without the support of any intelligent auxiliary system.

[0150] Comparative Example 2: Method based on a single deep learning model

[0151] This comparison adopts an end-to-end deep learning model to predict the puncture path directly from ultrasound images. Specifically, we use an improved U-Net model, which uses ResNet-50 as the backbone network in the encoder part to improve feature extraction capabilities.

[0152] Comparative Example 3: Method based on traditional image processing and search algorithm

[0153] This example combines traditional image processing techniques with path search algorithms. We use an adaptive threshold segmentation method to identify targets and obstacles, and then use the A* algorithm to plan the path.

[0154] In order to comprehensively evaluate the performance of these methods, we designed the following key indicators:

[0155] 1. Path planning accuracy: measures whether the planned path reaches the target location accurately. We use Euclidean distance to calculate the distance between the end point of the planned path and the actual target point. If the distance is less than 5mm, the planning is considered accurate.

[0156] 2. Safety index: Evaluate the minimum distance between the planned path and key structures (such as large blood vessels, nerves, etc.). We calculate the distance from all points on the path to the nearest key structure and take the minimum value as the safety index.

[0157] 3. Path smoothness: measured using the average curvature of the path. The higher the smoothness, the easier the path is to operate.

[0158] 4. Computation time: The total time from input image to output planned path, reflecting the real-time performance of the method.

[0159] 5. Risk assessment accuracy: Compare the system’s risk assessment results with the expert scores and calculate their correlation coefficient.

[0160] We tested 100 real liver puncture cases, and each method was run under the same hardware environment. The test results are shown in the following table:

[0161] index Example 1 Comparative Example 1 Comparative Example 2 Comparative Example 3 Path planning accuracy 95% 78% 87% 82% Safety index (mm) 8.7 5.2 7.1 6.5 Path Smoothness 0.92 0.75 0.83 0.79 Calculation time(s) 0.8 N / A 1.2 2.5 Risk assessment accuracy 0.91 0.62 0.78 0.73

[0162] From these results, we can clearly see that the method of the present invention (Example 1) performs well in all indicators and is significantly better than other comparative methods.

[0163] First, in terms of path planning accuracy, the method of the present invention achieved a high accuracy of 95%, far exceeding other methods. This is mainly due to the multi-scale deep learning segmentation and path planning algorithm based on deep reinforcement learning that we adopted. These advanced algorithms can better understand complex anatomical structures and find the optimal balance between multiple goals.

[0164] In terms of safety index, the method of the present invention also performed outstandingly, with an average safety distance of 8.7 mm. This reflects the effectiveness of our dynamic risk assessment mechanism, which can assess path risks in real time and make adjustments, greatly improving the safety of surgery.

[0165] The high score of path smoothness (0.92) shows that our method can not only find a safe path but also ensure the operability of the path. This is crucial for actual clinical operation because a smooth path is easier for doctors to execute accurately.

[0166] In terms of computing time, although our method involves complex deep learning and reinforcement learning algorithms, through optimized hardware configuration and parallel computing technology, we have controlled the computing time to 0.8 seconds, which fully meets the needs of real-time operation. This is especially important because in a dynamic surgical environment, fast response capability is directly related to the success rate of surgery.

[0167] Finally, in terms of risk assessment accuracy, our method also showed significant advantages, with a correlation coefficient of 0.91 with expert scores. This means that our system can provide risk assessments comparable to those of experts, providing a reliable reference for doctors' decision-making.

[0168] Please refer to Figure 8 , the multi-dimensional dynamic decision tree diagram shows the complexity and accuracy of real-time decision-making during the puncture process. The following is an explanation of the figure: The main path represents the overall path planning from the starting point to the end point of the puncture. This main path shows the key decision nodes in the whole process, and each node is an important decision point. The size of the node reflects the uncertainty at the decision point. Larger nodes indicate that there is greater uncertainty at the decision point, which means that the decision may be more complex or have more variables in actual operation. The color of the node indicates safety, from green (safe) to red (dangerous). The closer the color is to red, the higher the decision risk of the node. The thickness of the branch indicates the priority of the path or the possibility of selection. Thicker branches mean that the path is more likely to be selected or more important during the operation. The color change of the branch also indicates safety. As the path advances from the starting point to the end point, the gradual change of color reflects the change in risk level. Example 1 has more decision nodes, a finer branch structure, and an overall greener main path in the figure, showing higher safety and accuracy. The comparative methods will show fewer decision nodes, fewer branch choices, or the color tends to be yellow or red in the decision tree, indicating that these methods are insufficient in safety and accuracy. The decision tree diagram intuitively shows the advantages and disadvantages of different methods in complex surgical procedures. Through the structure of the decision tree, the size and color of the nodes, and the thickness and safety distribution of the path, it can be clearly seen that the method of the present invention has obvious advantages in the precision, safety and reliability of path planning.

[0169] In general, the multi-dimensional dynamic decision tree diagram not only shows the complexity of decision-making in each step of the surgical process, but also intuitively reflects the differences in safety, accuracy and operability of different methods. This provides an important reference for evaluating and selecting the best surgical path.

[0170] Please refer to Fig. 9 , this 3D puncture trajectory comparison chart shows the comparison of puncture paths of four different methods in a complex anatomical environment. The following is a description of the figure: Example 1 (the green trajectory represents the method proposed by the present invention. This path is relatively thin, indicating that the positioning of the puncture path is very precise and avoids important anatomical structures. The green color also shows that the path is relatively safe. Comparative Example 1, Comparative Example 2, Comparative Example 3 (blue, orange and red trajectories), these trajectories represent the comparison of other methods. The thickness and color changes of the path show different degrees of positioning uncertainty and safety. For example, the red trajectory indicates that this method has a higher risk in certain areas, while the thickness of the path shows that the positioning is not as accurate as Example 1. Important blood vessels (marked in purple) and target lesions (marked in black), the figure marks the key anatomical structures that require special attention during the puncture process. The puncture path of Example 1 obviously avoids important blood vessels, This shows that this method has obvious advantages in safety. At the same time, the path eventually reaches the target lesion, indicating that its target achievement rate is high. The trajectory of Example 1 not only avoids high-risk areas, but also shows high path smoothness and safety, which is very important in actual clinical operations. A smoother path means that the surgical operation is easier and safer. It can be seen from the thickness of the path that the path of Example 1 has higher accuracy and smaller positioning error, while the path of the comparative method shows greater uncertainty and risk. This figure intuitively shows the superiority of the method of the present invention in complex anatomical structures, including higher safety, more accurate path planning, and better clinical applicability. Through these comparisons, it can be clearly seen that this method is superior to traditional or other comparative methods in all indicators.

[0171] In general, these experimental results fully demonstrate the superiority of our method. It not only performs well in terms of accuracy and safety, but also provides high-quality path planning and risk assessment while ensuring real-time performance. This comprehensive performance advantage makes our method particularly suitable for complex minimally invasive surgical scenarios, such as targeted puncture treatment of liver tumors.

[0172] It is worth noting that although Comparative Example 2 (the method based on a single deep learning model) performs well in some indicators, it still has obvious deficiencies in risk assessment and path smoothness. This shows that relying solely on a single deep learning model is not enough, and it is necessary to consider multiple factors as our method does and introduce a more complex algorithm framework.

[0173] In summary, these experimental results not only verify the effectiveness of the method of the present invention, but also point out the direction for the development of future intelligent surgical planning systems. We believe that with the further improvement and promotion of this technology, it will bring revolutionary changes to minimally invasive surgery and ultimately benefit more patients.

[0174] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent puncture path planning method under ultrasound guidance, characterized in that: The following steps are involved: Acquire real-time ultrasound image sequences; Preprocessing and enhancing the real-time ultrasound image sequence to obtain an enhanced ultrasound image; Based on the enhanced ultrasound image, multi-scale deep learning image segmentation is performed to obtain a segmented anatomical structure label map; based on the segmented anatomical structure label map, target positioning and feature extraction are performed to obtain target position coordinates and feature descriptors; based on the segmented anatomical structure label map and the target position coordinates, deep reinforcement learning is used to perform path planning to obtain an optimal puncture path; based on the optimal puncture path and the enhanced ultrasound image, dynamic risk assessment is performed to obtain a risk assessment result; An augmented reality guidance interface is generated based on the optimal puncture path, the risk assessment result and the enhanced ultrasound image.

2. The method according to claim 1, characterized in that The step of preprocessing and enhancing the real-time ultrasonic image sequence includes: calculating the local standard deviation and the global standard deviation according to the local area of ​​each pixel point in the real-time ultrasonic image sequence; determining the adaptive filtering window size based on the ratio of the local standard deviation to the global standard deviation; using the adaptive filtering window size, performing median filtering on the real-time ultrasonic image sequence to obtain a denoised image sequence; calculating a local histogram for each image in the denoised image sequence; determining an adaptive clipping threshold according to the maximum and minimum values ​​of the local histogram; using the adaptive clipping threshold, performing contrast-limited adaptive histogram equalization on the denoised image sequence to obtain an equalized image sequence; calculating the local mean and the global mean of each image in the equalized image sequence; determining an adaptive Gamma correction parameter based on the difference between the local mean and the global mean; using the adaptive Gamma correction parameter, performing Gamma correction on the equalized image sequence to obtain the enhanced ultrasonic image.

3. The method according to claim 1, characterized in that The step of performing multi-scale deep learning image segmentation based on the enhanced ultrasound image includes: constructing a multi-scale U-Net++ network model, including an encoder and a decoder; in the encoder, extracting the multi-scale features of the enhanced ultrasound image through continuous convolution layers and pooling layers; in the decoder, gradually restoring the spatial resolution of the feature map through transposed convolution layers; adding dense jump connections between the encoder and the decoder to achieve the fusion of features of different scales; calculating the spatial attention map of the fused feature map, which reflects the importance of different areas of the image; calculating the channel attention vector of the fused feature map, which reflects the importance of different feature channels; multiplying the spatial attention map and the channel attention vector with the fused feature map to obtain a weighted feature map; based on the weighted feature map, generating a multi-category segmentation probability map through a 1x1 convolution layer and a softmax activation function; calculating Dice loss, boundary loss and Focal loss, and weightedly combining these three losses to obtain a total loss function; optimizing the parameters of the network model through back propagation and gradient descent, and finally obtaining the segmented anatomical structure label map.

4. The method according to claim 1, characterized in that: The step of performing target positioning and feature extraction based on the segmented anatomical structure label map includes: obtaining the segmented anatomical structure label maps of the current frame and the previous k frames; performing temporal consistency analysis on the k+1 frame label map to screen out stable connected areas; calculating the area, perimeter and shape complexity of the stable connected areas, and screening out candidate target areas according to a preset threshold; For pixels within the candidate target area, weights are calculated according to their grayscale values ​​in the enhanced ultrasound image; using the weights, the weighted centroid of the candidate target area is calculated to obtain the target position coordinates; taking the target position coordinates as the center, a local area of ​​a fixed size is extracted; shape features of the local area are calculated, including roundness, eccentricity and directionality; texture features of the local area are calculated, including grayscale co-occurrence matrix features and local binary pattern features; using a pre-trained convolutional neural network, deep features of the local area are extracted as context features; the shape features, the texture features and the context features are spliced ​​to form the feature descriptor.

5. The method according to claim 1, characterized in that The step of using deep reinforcement learning to perform path planning based on the segmented anatomical structure label map and the target position coordinates includes: calculating the distance from each position to the nearest obstacle according to the segmented anatomical structure label map to obtain an obstacle distance map; generating an uncertainty map based on the confidence of the segmented anatomical structure label map; estimating the tissue deformation field by using the registration results between consecutive frames to obtain a deformation degree map; calculating a local vascular density map according to the vascular segmentation result; weightedly combining the obstacle distance map, the uncertainty map, the deformation degree map and the vascular density map to obtain a comprehensive risk map; constructing a reinforcement learning environment based on the comprehensive risk map and the target position coordinates; and in the reinforcement learning environment, using an improved Deep The Q-Network algorithm is used for global path planning; based on the global path, the newly designed AdvantageActor-Critic algorithm is used for local path optimization; an adaptive reward function is designed according to the safety, efficiency and accuracy of the path; multiple agents are introduced, each of which is responsible for different subtasks, such as safety, efficiency and accuracy; a knowledge distillation mechanism is used to enable knowledge sharing and collaborative learning among the agents; an integrated teacher strategy is constructed based on the strategies of all agents; the integrated teacher strategy is used to guide the learning process of each agent; combined with the learned Q value, an improved Rapidly-exploring Random Trees algorithm is used for path sampling and optimization; iterative optimization is performed until convergence or a preset number of iterations is reached to obtain the optimal puncture path.

6. The method according to claim 5, characterized in that The steps of constructing the dynamic risk map include: initializing weight coefficients based on the obstacle distance map, the uncertainty map, the deformation degree map and the blood vessel density map; calculating the difference between the current risk map and the target risk distribution; updating the weight coefficients based on the difference using the gradient descent method; recalculating the comprehensive risk value using the updated weight coefficients; repeating the above steps until the risk map converges or reaches a preset number of iterations; and outputting the final dynamic risk map.

7. The method according to claim 1, characterized in that The steps of performing dynamic risk assessment based on the optimal puncture path and the enhanced ultrasound image include: calculating the path curvature along the optimal puncture path; determining the sampling interval based on the path curvature and the dynamic risk map; extracting path key points according to the sampling interval; extracting obstacle distance, uncertainty, tissue deformation and vascular density information of the local area for each path key point; converting the extracted information into a fuzzy set using fuzzy logic rules; calculating the basic probability distribution of each risk factor based on evidence theory; fusing the evidence of multiple risk factors through the Dempster combination rule to obtain a local risk assessment result; constructing a recursive neural network and taking the local risk assessment result sequence as input; extracting the global features of the optimal puncture path; splicing the global features with the hidden state of the recursive neural network; and outputting the overall path risk assessment result through a fully connected layer and a Sigmoid activation function.

8. The method according to claim 1, characterized in that The step of generating an augmented reality guidance interface based on the optimal puncture path, risk assessment results and enhanced ultrasound images includes: calculating a saliency map and a contrast map of the enhanced ultrasound image; generating an adaptive fusion weight map based on the saliency map and the contrast map; using the adaptive fusion weight map, superimposing the optimal puncture path on the enhanced ultrasound image; calculating the global mean and standard deviation of the dynamic risk map; determining an adaptive risk threshold based on the global mean and the standard deviation; marking high-risk areas on the superimposed image based on the adaptive risk threshold; constructing a Kalman filter to initialize needle tip state estimation; extracting a needle tip feature template in the enhanced ultrasound image; using the normalized cross-correlation method to perform template matching in the predicted area; updating the needle tip position estimation by combining the Kalman filter prediction and template matching results; calculating the distance from the needle tip to the target, the current local risk value and the needle tip movement speed; dynamically adjusting the color and transparency of the visual feedback based on the above calculation results; and at the same time, adjusting the frequency and volume of the audio feedback based on the above calculation results; synthesizing all visual elements into a single layer to generate a final augmented reality guidance interface.

9. The method according to any one of claims 1 to 8, characterized in that: Before executing the method, the following data preparation steps are also included: obtaining the model and parameter information of the ultrasound equipment to ensure that its frame rate is not less than 30fps and the resolution is not less than 1024x768; collecting relevant medical information of the patient, including age, gender, medical history and relevant examination results; obtaining the surgical plan, including puncture target, entry point selection and avoidance structure; loading pre-trained deep learning model parameters, including image segmentation model and feature extraction model; Initialize the policy network and value network parameters of the reinforcement learning agent; calibrate the spatial position and orientation of the ultrasound probe; perform a system self-check to ensure that all modules are working properly.

10. An intelligent puncture path planning system under ultrasound guidance for implementing the method according to any one of claims 1 to 9, characterized in that: include: An image acquisition module, used for obtaining a real-time ultrasound image sequence; An image preprocessing module, used for preprocessing and enhancing the real-time ultrasound image sequence; An image segmentation module, used for performing multi-scale deep learning image segmentation on the enhanced ultrasound image; Target localization module, used for target localization and feature extraction based on segmentation results; A path planning module is used for path planning using deep reinforcement learning; a risk assessment module is used for dynamic risk assessment; a visual guidance module is used for generating an augmented reality guidance interface; wherein the image acquisition module includes an ultrasound probe and an image acquisition card; the image preprocessing module, image segmentation module, target positioning module, path planning module and risk assessment module are all implemented by a computer with GPU acceleration function; the visual guidance module includes a display and a stereo sound system.

Citation Information

Patent Citations

  • Liver tumor image enhancement and segmentation method based on deep learning

    CN115578404A

  • Puncture needle ultrasound image segmentation method, device and system

    CN116152266A

  • Deep venipuncture path generation method and device

    CN116704213A

  • Medical image recognition method and system based on enhanced annotation and deep learning

    CN118096721A

  • Ultrasonic navigation method and system for puncture robot

    CN118285919A

Cited By

  • Infrared and visible light image fusion method for inspection of power transmission and transformation equipment

    CN119540702A

  • Tumor ablation puncture operation planning method and system

    CN119950025A

  • Tumor ablation puncture surgery planning method and system

    CN119950025B

  • Path marking and automatic planning method for individualized virtual laparoscopic surgical robot

    CN120002627A

  • Ultrasonic guided pelvis puncture method based on deep learning

    CN120053035A