Positioning puncture surgery robot system with reconfigurable working space

Abstract

The invention discloses a positioning puncture surgery robot system with a reconfigurable working space, and the system comprises a multi-mode image collection module which is used for obtaining a high-resolution image of a target region of a patient; the three-dimensional anatomical model reconstruction module is used for generating a three-dimensional anatomical model containing a skeleton structure and a soft tissue structure; the target point selection and puncture path generation module is used for automatically recommending puncture target points according to operation types and clinical requirements and automatically generating a plurality of preliminary puncture paths; the multi-plane segmentation and path injection module is used for segmenting the three-dimensional anatomical model from a coronal plane, a sagittal plane and an axial plane and generating slices, and injecting puncture paths into the corresponding slices; by integrating multi-modal image acquisition, three-dimensional anatomical model reconstruction and automatic path planning and optimization setting, the operation efficiency and accuracy are remarkably improved, the operation risk is reduced through real-time data analysis, and a more personalized treatment scheme is provided for a patient.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a positioning puncture surgical robot system with a reconfigurable working space. Background Art

[0002] In the field of medical technology, puncture surgery is a common interventional treatment method, which is widely used in biopsy, drainage and treatment scenarios. However, traditional methods have significant limitations in image data acquisition and processing, three-dimensional anatomical model reconstruction, and puncture path planning. First, single-modality imaging (such as CT or MRI) cannot provide detailed information on hard and soft tissues at the same time, which limits the comprehensiveness and accuracy of surgical planning. Secondly, there may be delays or losses in the transmission of image data from the imaging device to the processing system, which affects the timeliness and accuracy of image data. In addition, manual segmentation of important structures (such as skull, cerebral cortex, nerves, and blood vessels) is not only time-consuming and labor-intensive, but also easily affected by subjective factors, which reduces the accuracy of the model. The precise spatial alignment of multimodal data is also a complex process. Traditional methods are difficult to ensure the accuracy and consistency of the registration, which affects the quality of the three-dimensional anatomical model.

[0003] There are also challenges in puncture path planning. Traditional methods mainly rely on the doctor's experience to select the puncture target point and design the path. There is a lack of automated tools to recommend the best plan based on the type of surgery and clinical needs, which increases the risk of surgery. Existing systems are insufficient in evaluating the safety of puncture paths. They cannot effectively calculate the minimum distance between points on the path and important structures, nor can they detect whether the path passes directly through these structures, which may cause damage to critical structures during surgery. In addition, traditional methods are limited to slice analysis of a single plane and cannot provide a comprehensive view of the coronal, sagittal, and axial planes, limiting doctors' comprehensive assessment of path safety and feasibility; the lack of the function of mapping the puncture path to different slices makes it difficult for doctors to intuitively evaluate the path design, increasing the difficulty of surgical operations.

[0004] Therefore, it is necessary to provide a positioning puncture surgical robot system with a reconfigurable workspace to solve the above technical problems. Summary of the invention

[0005] The present invention overcomes the deficiencies of the prior art and provides a positioning puncture surgical robot system with a reconfigurable workspace.

[0006] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a positioning puncture surgical robot system with reconfigurable workspace, comprising:

[0007] A multimodal image acquisition module for obtaining high-resolution images of the patient's target area;

[0008] A three-dimensional anatomical model reconstruction module is used to generate a three-dimensional anatomical model including bone structure and soft tissue structure;

[0009] The target point selection and puncture path generation module is used to automatically recommend puncture target points and automatically generate several preliminary puncture paths according to the type of surgery and clinical needs;

[0010] The multi-plane segmentation and path injection module is used to segment the three-dimensional anatomical model from the coronal plane, sagittal plane and axial plane and generate slices, and inject the puncture path into the corresponding slices;

[0011] A texture feature extraction and analysis module is used to automatically extract texture features in slices and identify important structures, including blood vessels, nerves, and brain tissues, to assess the safety of the puncture path;

[0012] The path screening and optimization module is used to score each path based on safety, trauma minimization, and puncture angle factors, and screen or adjust the optimal puncture path;

[0013] The texture feature extraction and analysis module extraction methods include but are not limited to gray level co-occurrence matrix, local binary pattern and gradient direction histogram, and the path screening and optimization module optimization process ensures that the path maintains a safe distance from important structures:

[0014] In a preferred embodiment of the present invention, the multimodal image acquisition module includes: a CT scanner, an MRI scanner and an image transmission interface.

[0015] In a preferred embodiment of the present invention, the three-dimensional anatomical model reconstruction module includes: an automatic segmentation algorithm, a multimodal registration algorithm and three-dimensional reconstruction software.

[0016] In a preferred embodiment of the present invention, the target point selection and puncture path generation module is used to perform subsequent analysis based on the doctor's manual selection, and automatically generate the puncture path by considering the starting point, the target point and possible turning points in the middle.

[0017] In a preferred embodiment of the present invention, the texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures, and detecting whether the path directly passes through important structures.

[0018] In a preferred embodiment of the present invention, a method for using a positioning puncture surgical robot system with a reconfigurable workspace comprises the following steps:

[0019] S1: Acquire high-resolution images of the patient's target area through a multimodal image acquisition module;

[0020] S2: Generate a three-dimensional anatomical model including bone structure and soft tissue structure using a three-dimensional anatomical model reconstruction module;

[0021] S3: Based on the target point selection and puncture path generation module, the puncture target points are automatically recommended according to the surgical type and clinical needs, and several preliminary puncture paths are automatically generated;

[0022] S4: The three-dimensional anatomical model is segmented and sliced ​​from the coronal, sagittal and axial planes through the multi-plane segmentation and path-injection module, and the puncture path is injected into the corresponding slices;

[0023] S5: Use the texture feature extraction and analysis module to automatically extract important structures of texture features in the slices and evaluate the safety of the puncture path;

[0024] S6: The path screening and optimization module is used to score each path based on safety, trauma minimization, and puncture angle factors, and to screen or adjust the optimal puncture path.

[0025] In a preferred embodiment of the present invention, in S1, the multimodal image acquisition module transmits the acquired image data to the computer processing system in real time to ensure the accuracy and timeliness of the image data.

[0026] In a preferred embodiment of the present invention, in S2, the automatic segmentation algorithm in the three-dimensional anatomical model reconstruction module is based on deep learning technology, which is used to automatically identify and segment important structures of the skull, cerebral cortex, nerves, and blood vessels, and provide a high-precision three-dimensional model.

[0027] In a preferred embodiment of the present invention, in S5, the texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures, and detecting whether the path directly passes through important structures.

[0028] In a preferred embodiment of the present invention, in S6, the path screening and optimization module comprehensively evaluates the safety, trauma minimization, and puncture angle factors of the path, thereby screening out one or more optimal puncture paths.

[0029] The present invention solves the defects existing in the background technology and has the following beneficial effects:

[0030] (1) The present invention provides a positioning puncture surgical robot system with a reconfigurable workspace. By integrating multimodal image acquisition, three-dimensional anatomical model reconstruction, automatic path planning and optimization settings, a highly automated, safe and reliable surgical assistance platform is constructed. The system not only significantly improves surgical efficiency and accuracy, but also reduces surgical risks through real-time data analysis and provides patients with more personalized treatment plans. In addition, the modular design of the system allows it to flexibly adjust the workspace according to different clinical needs, enhancing adaptability and practicality.

[0031] (2) The present invention provides a positioning puncture surgical robot system with a reconfigurable workspace, which uses a multimodal image acquisition module to obtain high-resolution images, combines deep learning technology to achieve automatic segmentation, and ensures data alignment under different imaging modes through a multimodal registration algorithm. This combination ensures a high degree of restoration of the patient's internal structure, making the generated three-dimensional anatomical model more accurate, laying the foundation for subsequent target point selection and puncture path planning. Doctors can make more accurate judgments based on more realistic anatomical information, reduce the risk caused by misjudgment, and improve the success rate of surgery. Compared with traditional single-modal imaging methods or manual modeling methods, the multimodal data fusion and automated processing provided by the present invention greatly improve the accuracy of the model and the safety of surgery. Traditional methods often rely on the doctor's experience to make judgments, which are subjective and uncertain, while the present invention achieves objective quantitative evaluation through algorithm optimization. It helps to shorten the preoperative preparation time, while improving the success rate of surgery and reducing the possibility of postoperative complications, thereby improving the patient's recovery speed and quality of life.

[0032] (3) The present invention provides a positioning puncture surgical robot system with a reconfigurable workspace, so that the target point selection and puncture path generation module can automatically recommend puncture target points and automatically generate several preliminary puncture paths according to the type of surgery and clinical needs; the path screening and optimization module considers multiple factors such as safety, trauma minimization, and puncture angle. Through a comprehensive evaluation of the safety and feasibility of different paths, the module can effectively avoid damage to important tissues and organs, such as blood vessels and nerves, while ensuring the selection of the optimal puncture angle. Ensure the selection of the best path, thereby minimizing harm to the patient and improving the safety and success rate of the operation. Traditional methods often rely on the doctor's experience for path planning, which has a certain degree of subjectivity and uncertainty. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 is a schematic diagram of the system structure of a preferred embodiment of the present invention;

[0035] Figure 2 It is a flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0038] like Figure 1 As shown, the present invention provides a positioning puncture surgical robot system with a reconfigurable workspace, comprising:

[0039] In the present invention, the multimodal image acquisition module is used to obtain high-resolution images of the target area of ​​the patient, and the multimodal image acquisition module includes: a CT scanner, an MRI scanner, and an image transmission interface;

[0040] In one embodiment, a CT scanner can quickly obtain cross-sectional images of structures in a patient's body, and is particularly good at showing bones and calcified tissues. The high-contrast images it provides are essential for determining the location and morphology of hard skull tissues.

[0041] MRI scanners are better at imaging soft tissue, brain tissue, blood vessels, and nerves than CT scanners. MRI provides rich soft tissue contrast, which helps to depict the internal structure of the brain and the specific location of blood vessels and nerves in detail.

[0042] The image transmission interface is the bridge between the imaging device and the computer processing system, ensuring that the image data of CT scanners and MRI scanners can be transmitted to the computer in a timely and accurate manner for further processing. This includes large-capacity data transmission through high-speed networks or dedicated hardware.

[0043] A three-dimensional anatomical model reconstruction module is used to generate a three-dimensional anatomical model including a bone structure and a soft tissue structure. The three-dimensional anatomical model reconstruction module includes: an automatic segmentation algorithm, a multimodal registration algorithm, and a three-dimensional reconstruction software;

[0044] In the present invention, the automatic segmentation algorithm is 3D U-Net, which is based on advanced deep learning technology and can intelligently identify and segment different anatomical structures, such as skull, cerebral cortex, nerves, and blood vessels. By training a large number of annotated medical image data sets, 3D U-Net can not only process hard tissues, but also segment soft tissues.

[0045] In one embodiment, the multimodal registration algorithm is B-spline MMR, which is used to accurately align the data from the CT scanner and the MRI scanner imaging mode. The three-dimensional reconstruction software is 3DSlicer: the three-dimensional reconstruction software uses advanced geometric modeling technology and surface reconstruction algorithms to construct a realistic three-dimensional anatomical model based on the segmented image data.

[0046] In the present invention, the target point selection and puncture path generation module is used to automatically recommend puncture target points and automatically generate several preliminary puncture paths according to the type of surgery and clinical needs, and is used to perform subsequent analysis based on the doctor's manual selection and automatically generate the puncture path by considering the starting point, target point and possible turning points in the middle.

[0047] In one embodiment, the target point selection and puncture path generation module can intelligently recommend the best puncture target point based on the type of surgery (such as ventricular puncture, tumor biopsy) and clinical needs (such as avoiding important structures), and combined with the three-dimensional anatomical model to ensure that the recommended target point is consistent with anatomical principles and takes into account individual differences among patients.

[0048] For example, in ventricular puncture, the system will recommend the anterior horn, posterior horn or basal ganglia of the lateral ventricle as common puncture sites. For each selected target point, the system automatically generates multiple puncture paths from the starting point (usually the skull surface or other easily accessible locations) to different combinations of target points, taking into account the path length, angle, and tissue type along the way.

[0049] In some cases, the path may need to include one or more turning points to avoid important nerves or blood vessels. The system can reasonably design the turning point positions on the path while maintaining a safe distance. In addition, doctors can manually select target points. Once a new target point is selected, the system will quickly update the path planning and provide real-time feedback on the safety and feasibility of the new path to help doctors make the best decision.

[0050] In the present invention, the multi-plane segmentation and path injection module is used to segment the three-dimensional anatomical model from the coronal plane, sagittal plane and axial plane and generate slices, and at the same time inject the puncture path into the corresponding slices.

[0051] In one embodiment, the multi-plane segmentation and path-injection module is responsible for segmenting the three-dimensional anatomical model from the coronal plane, sagittal plane and axial plane, and generating corresponding slice images. In addition, the multi-plane segmentation and path-injection module also maps the puncture path into two-dimensional slices of the coronal plane, sagittal plane and axial plane, so as to more intuitively evaluate the safety and feasibility of the path.

[0052] Multi-plane segmentation: Receive the 3D anatomical model from the 3D anatomical model reconstruction module, which includes the skull, cerebral cortex, nerves, and blood vessels. Set the slice thickness to 1-2mm to obtain slice images of different resolutions. Thinner slices can provide fine anatomical details, but the calculation is larger; thicker slices can speed up the analysis and are suitable for preliminary screening.

[0053] When segmenting the coronal plane: The coronal plane refers to the vertical plane that passes through the human body from front to back. In the three-dimensional coordinate system, the coronal plane is parallel to the XZ plane, that is, the Y-axis coordinate value is fixed. Along the fixed Y-axis coordinate value, the slice position is gradually moved to generate a series of coronal slices. Each slice represents the two-dimensional projection of the three-dimensional model on the coronal plane. By adjusting the increment of the Y-axis coordinate value, the distance between slices is controlled, thereby determining the slice thickness.

[0054] When segmenting the sagittal plane: The sagittal plane refers to the vertical plane that passes through the human body from left to right. In the three-dimensional coordinate system, the sagittal plane is parallel to the YZ plane, that is, the X-axis coordinate value is fixed. Along the fixed X-axis coordinate value, the slice position is gradually moved to generate a series of sagittal slices. Each slice represents the two-dimensional projection of the three-dimensional model on the sagittal plane. By adjusting the increment of the X-axis coordinate value, the distance between slices is controlled, thereby determining the slice thickness.

[0055] When segmenting the axial plane: The axial plane refers to the horizontal plane that passes through the human body from top to bottom. In the three-dimensional coordinate system, the axial plane is parallel to the XY plane, that is, the Z-axis coordinate value is fixed. Along the fixed Z-axis coordinate value, the slice position is gradually moved to generate a series of axial slices. Each slice represents the two-dimensional projection of the three-dimensional model on the axial plane. By adjusting the increment of the Z-axis coordinate value, the distance between slices is controlled, thereby determining the slice thickness.

[0056] Path Emission Process:

[0057] The key point data of each puncture path is obtained from the target point selection and puncture path generation module. The key points include the starting point, the target point, and any turning points on the path. The position information of each point on the path (such as X, Y, Z coordinates) is converted into a continuous parameter form.

[0058] Determine the location of the waypoints on each plane:

[0059] Coronal plane projection: Keep the Y-axis coordinate unchanged and gradually move the position on the XZ plane to generate a series of coronal slices. For each slice, find all points on the path that meet the Y-axis coordinate. For each path point that meets the conditions, project it onto the corresponding coronal slice to form a two-dimensional coordinate (X, Z).

[0060] Sagittal plane projection: Keep the X-axis coordinate unchanged and gradually move the position on the YZ plane to generate a series of sagittal slices. For each slice, find all points on the path that meet the X-axis coordinate. For each path point that meets the conditions, project it onto the corresponding sagittal slice to form a two-dimensional coordinate (Y, Z).

[0061] Axial plane projection: Keep the Z-axis coordinate unchanged and gradually move the position on the XY plane to generate a series of axial slices. For each slice, find all points on the path that meet the Z-axis coordinate. For each path point that meets the conditions, project it onto the corresponding axial slice to form a two-dimensional coordinate (X, Y).

[0062] When connecting path segments, first ensure the continuity of path segments between adjacent slices. By checking and adjusting the same path points that fall on multiple adjacent slices, they are also kept coherent in the two-dimensional slices. Based on the projection results of the path points, adjacent path points are connected with line segments on each slice to form a complete path segment. This allows doctors to review the path from multiple angles to ensure the rationality and safety of the path. When the doctor adjusts the puncture path, the system recalculates the new path projection in real time and updates the display. For example, if the doctor changes the starting point of the path or adds a new turning point, the module immediately reflects these changes in all relevant slices to keep the information consistent and up-to-date.

[0063] In the present invention, the texture feature extraction and analysis module is used to automatically extract texture features in slices and identify important structures, including blood vessels, nerves and brain tissues, to evaluate the safety of the puncture path. The texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures, and detecting whether the path directly passes through important structures. The texture feature extraction and analysis module extraction method includes but is not limited to grayscale co-occurrence matrix, local binary pattern and gradient direction histogram.

[0064] In one embodiment, the texture feature extraction and analysis module identifies blood vessels, nerves, and brain tissues through automatic processing of slice images, and evaluates the safety of the puncture path. The texture feature extraction and analysis module marks the safety of the path by calculating the minimum distance between points on the path and important structures and detecting whether the path directly passes through these structures.

[0065] Texture feature extraction methods include:

[0066] Gray-level co-occurrence matrix: used to describe the gray-level distribution relationship between pixels in an image and reflect the texture characteristics of the tissue;

[0067] The application steps of the gray level co-occurrence matrix are as follows: for each pixel in each slice, calculate the gray level relationship between it and the surrounding pixels and construct a gray level co-occurrence matrix. The gray level co-occurrence matrix records the frequency of occurrence of different gray level combinations;

[0068] Extract statistics of contrast, homogeneity, energy, and entropy from the gray-level co-occurrence matrix. These statistics reflect the texture characteristics of the image region and help distinguish different tissue types;

[0069] For example, blood vessels and nerves often have unique texture properties that can be identified through statistics of contrast, homogeneity, energy, and entropy.

[0070] Local binary pattern: used to capture local texture information in an image, especially suitable for identifying the edges of blood vessels and nerves;

[0071] The application steps of local binary pattern are as follows: Calculate the local binary pattern code for each pixel, that is, compare the central pixel with the pixels in its neighborhood, and generate a binary number to represent the local texture information of the position. For example, if the central pixel is brighter than a neighboring pixel, the corresponding position is recorded as 1, otherwise it is recorded as 0;

[0072] The LBP codes of all pixels are counted to form a histogram. Different tissue types usually have unique LBP histogram distributions, so specific structures can be identified by matching histograms;

[0073] For example, the edges of blood vessels often appear as high-contrast local binary patterns, while brain tissue may show a more uniform pattern.

[0074] Gradient orientation histogram: used to describe the shape and orientation of objects in an image, which helps to distinguish different types of tissues;

[0075] The application steps of the gradient direction histogram are as follows: Calculate the gradient size and direction at each pixel. Gradient information can capture edges and contours. For example, blood vessels usually show strong gradient changes. Divide the image into small blocks and calculate the gradient direction histogram in each block. Finally, summarize the histograms of each block to form the gradient direction histogram feature vector of the entire image. This localized processing method enhances the ability to describe complex textures and shapes.

[0076] The features extracted by gray-level co-occurrence matrix, local binary pattern and gradient orientation histogram are combined to construct a multi-dimensional feature vector to comprehensively describe the texture information of each pixel and its neighborhood.

[0077] Using deep learning frameworks (such as convolutional neural networks), classifiers are trained based on a large number of annotated medical image datasets. These classifiers can accurately distinguish important structures such as blood vessels, nerves, and brain tissue based on the extracted texture features.

[0078] When a new slice image is input, the system will apply the trained classifier to automatically identify and mark the important structures in the image. For example, blood vessels may be marked in red, nerves in blue, and brain tissue in green, making it easier for doctors to intuitively assess the safety of the puncture path.

[0079] During the safety assessment, for each point on each puncture path, the Euclidean distance between it and the nearest important structure is calculated. If this distance is less than the preset safety threshold (such as 1-2mm), the path is considered to have a safety hazard and the corresponding position on the path is marked. In addition to the distance calculation, the system also checks whether the path passes directly through important blood vessels or nerves. This detection is based on the probability mapping of paths and structures. If a point on the path is found to have a high probability of belonging to a certain important structure, it is immediately marked as "unsafe".

[0080] In the present invention, the path screening and optimization module is used to score each path according to safety, trauma minimization, and puncture angle factors, and screen or adjust the optimal puncture path. The optimization process of the path screening and optimization module ensures that the path maintains a safe distance from important structures.

[0081] In one embodiment:

[0082] 1. Define the scoring criteria

[0083] Safety: Evaluates the minimum distance between the path and important structures and whether it passes directly through them.

[0084] Minimize trauma: Consider the path length, the type and amount of tissue passed through, and minimize damage to surrounding tissues.

[0085] Penetration angle: The angle of the evaluation path relative to the target point, ideally should be close to vertical entry.

[0086] 2. Quantify the factors

[0087] Safety Rating:

[0088] Minimum distance: The point on the path is scored based on the distance from the nearest important structure. The larger the distance, the higher the score.

[0089] Crossing detection: If the path passes directly through an important structure, the safety score of the path is 0.

[0090] Trauma minimization score:

[0091] Path length: Shorter paths score higher.

[0092] Tissue type: Pathways that pass through less sensitive tissues score higher.

[0093] Number of organizations: The fewer organizations a path passes through, the higher the score.

[0094] Puncture Angle Scoring:

[0095] Ideal angle: The highest score is obtained when the path angle is close to 90 degrees, and the greater the deviation angle, the lower the score.

[0096] 3. Comprehensive Rating

[0097] The composite score is calculated using the weighted sum formula:

[0098] Comprehensive score = ω1 × safety score + ω2 × trauma minimization score + ω3 × puncture angle score, where ω1, ω2 and ω3 are the weights of each factor, satisfying ω1+ω2+ω3=1.

[0099] 4. Path screening and adjustment

[0100] Filter the optimal path: Based on the comprehensive scoring results, select the path with the highest score as the optimal choice.

[0101] Interactive Adjustment: Provides the function for doctors to manually adjust the pathway, and updates the score in real time after each adjustment to ensure that the new pathway meets all optimization criteria.

[0102] 5. Ensure a safe distance

[0103] Throughout the process, the distance between the path and important structures is monitored to ensure that the path maintains an adequate safety distance, and any changes that may result in insufficient safety distance are corrected immediately.

[0104] 6. Decision support

[0105] Ultimately, one or more optimal pathways are recommended and a detailed scoring report is provided to help doctors make informed decisions.

[0106] like Figure 2 As shown, a method for using a positioning puncture surgical robot system with a reconfigurable workspace includes the following steps:

[0107] S1: A high-resolution image of the patient's target area is obtained through a multimodal image acquisition module, which transmits the obtained image data to a computer processing system in real time to ensure the accuracy and timeliness of the image data;

[0108] S2: Generate a 3D anatomical model including bone structure and soft tissue structure using the 3D anatomical model reconstruction module. The automatic segmentation algorithm in the 3D anatomical model reconstruction module is based on deep learning technology and is used to automatically identify and segment important structures to provide a high-precision 3D model.

[0109] S3: Based on the target point selection and puncture path generation module, the puncture target points are automatically recommended according to the surgical type and clinical needs, and several preliminary puncture paths are automatically generated;

[0110] S4: The three-dimensional anatomical model is segmented and sliced ​​from the coronal, sagittal and axial planes through the multi-plane segmentation and path-injection module, and the puncture path is injected into the corresponding slices;

[0111] S5: Use the texture feature extraction and analysis module to automatically extract important structures of texture features in the slices and evaluate the safety of the puncture path. The texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures and detecting whether the path directly passes through important structures;

[0112] S6: The path screening and optimization module is used to score each path according to safety, trauma minimization, and puncture angle factors, and to screen or adjust the optimal puncture path. The path screening and optimization module conducts a comprehensive evaluation based on the safety, trauma minimization, and puncture angle factors of the path, thereby screening out one or more optimal puncture paths.

[0113] Example 1: Application in ventriculostomy

[0114] S1: A CT scanner is used to obtain cross-sectional images of the skull and internal structures. CT scans are fast and are particularly useful for showing bones and calcified tissue. The patient then enters an MRI scanner to obtain detailed images of the brain, blood vessels, and nerves. All acquired images are transmitted to a computer processing system via a high-speed network within the hospital or dedicated hardware;

[0115] S2: The 3D U-Net algorithm is based on deep learning technology and can intelligently identify and segment different anatomical structures, including the skull, cerebral cortex, nerves, and blood vessels. The B-spline MMR algorithm is used to accurately align data from different imaging modes of CT and MRI in space, so that the two imaging data can be fused in the same coordinate system to form a complete three-dimensional anatomical model. The 3D Slicer software uses advanced geometric modeling technology and surface reconstruction algorithms to construct realistic three-dimensional anatomical models based on segmented image data;

[0116] S3: In ventricular puncture, combined with the three-dimensional anatomical model, the system will recommend the anterior horn of the lateral ventricle as a common puncture site; for the selected target point, the system automatically generates multiple possible puncture paths. Each path goes from the starting point (usually the skull surface or other easily accessible location) to the target point, taking into account factors such as path length, angle, and tissue type along the way. If the path needs to avoid important nerves or blood vessels, the turning point position on the path will be reasonably designed;

[0117] S4: The 3D anatomical model was segmented according to the coronal, sagittal, and axial planes, and the slice thickness was set to 1.5 mm to obtain slice images of different resolutions. The key points of the puncture path (such as the starting point, target point, and any turning points) were converted into two-dimensional coordinates and mapped to the corresponding slices of the coronal, sagittal, and axial planes.

[0118] S5: Using the gray-level co-occurrence matrix, local binary pattern and gradient direction histogram methods, texture features are automatically extracted in each slice to identify blood vessels, nerves and brain tissues. The safety of the path is marked by calculating the minimum distance between points on the path and blood vessels, nerves and brain tissues, and detecting whether the path directly passes through these structures. If the distance is less than the preset safety threshold of 1mm, or the path directly passes through blood vessels, nerves and brain tissues, the path is considered to have a safety hazard and the corresponding position on the path is marked.

[0119] S6: Path screening and optimization

[0120] Scoring criteria definition:

[0121] Safety: Evaluates the minimum distance between the path and important structures and whether it passes directly through them.

[0122] Minimize trauma: Consider the path length, the type and amount of tissues passed through, and minimize damage to surrounding tissues;

[0123] Puncture angle: The angle of the evaluation path relative to the target point should ideally be close to vertical entry;

[0124] Quantify the factors:

[0125] Safety score: score is given based on the minimum distance, and a score of 0 is given if it directly crosses important structures;

[0126] Trauma minimization score: Shorter paths have higher scores, and paths that pass through less sensitive tissue and a smaller number of tissues also have higher scores;

[0127] Puncture angle score: The highest score is obtained when the path angle is close to 90 degrees;

[0128] Comprehensive score: The comprehensive score is calculated using a weighted sum formula, and the weight distribution can be adjusted according to actual needs;

[0129] Path screening and adjustment: Based on the comprehensive scoring results, the path with the highest score is selected as the optimal choice. At the same time, the system supports doctors to manually adjust the path, and the score is updated in real time after each adjustment to ensure that the new path meets all optimization criteria. During the whole process, the distance between the path and important structures is monitored to ensure that the path maintains a sufficient safety distance.

[0130] The above is based on the ideal embodiment of the present invention. Through the above description, relevant personnel can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the content in the specification, but must be determined according to the scope of the claims.

Claims

1. A positioning puncture surgical robot system with reconfigurable workspace, characterized in that: include: A multimodal image acquisition module for obtaining high-resolution images of the patient's target area; A three-dimensional anatomical model reconstruction module is used to generate a three-dimensional anatomical model including bone structure and soft tissue structure; The target point selection and puncture path generation module is used to automatically recommend puncture target points and automatically generate several preliminary puncture paths according to the type of surgery and clinical needs; The multi-plane segmentation and path injection module is used to segment the three-dimensional anatomical model from the coronal plane, sagittal plane and axial plane and generate slices, and inject the puncture path into the corresponding slices; A texture feature extraction and analysis module is used to automatically extract texture features in slices and identify important structures, including blood vessels, nerves, and brain tissues, to assess the safety of the puncture path; The path screening and optimization module is used to score each path based on safety, trauma minimization, and puncture angle factors, and screen or adjust the optimal puncture path; Among them, the extraction method of the texture feature extraction and analysis module includes but is not limited to gray level co-occurrence matrix, local binary pattern and gradient direction histogram, and the optimization process of the path screening and optimization module ensures that the path maintains a safe distance from important structures.

2. The positioning puncture surgical robot system with reconfigurable workspace according to claim 1, characterized in that: The multimodal image acquisition module includes: a CT scanner, an MRI scanner and an image transmission interface.

3. The positioning puncture surgical robot system with reconfigurable workspace according to claim 1, characterized in that: The three-dimensional anatomical model reconstruction module includes: an automatic segmentation algorithm, a multi-modal registration algorithm and three-dimensional reconstruction software.

4. The positioning puncture surgical robot system with reconfigurable workspace according to claim 1, characterized in that: The target point selection and puncture path generation module is used to perform subsequent analysis based on the doctor's manual selection and automatically generate the puncture path by considering the starting point, target point and possible turning points in the middle.

5. The positioning puncture surgical robot system with reconfigurable workspace according to claim 1, characterized in that: The texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures, and detecting whether the path directly passes through important structures.

6. A method for using a positioning puncture surgical robot system with reconfigurable workspace according to any one of claims 1 to 5, characterized in that: The following steps are involved: S1: Acquire high-resolution images of the patient's target area through a multimodal image acquisition module; S2: Generate a three-dimensional anatomical model including bone structure and soft tissue structure using a three-dimensional anatomical model reconstruction module; S3: Based on the target point selection and puncture path generation module, the puncture target points are automatically recommended according to the surgical type and clinical needs, and several preliminary puncture paths are automatically generated; S4: The three-dimensional anatomical model is segmented and sliced ​​from the coronal, sagittal and axial planes through the multi-plane segmentation and path-injection module, and the puncture path is injected into the corresponding slices; S5: Use the texture feature extraction and analysis module to automatically extract important structures of texture features in the slices and evaluate the safety of the puncture path; S6: The path screening and optimization module is used to score each path based on safety, trauma minimization, and puncture angle factors, and to screen or adjust the optimal puncture path.

7. The method for using the positioning puncture surgical robot system with reconfigurable workspace according to claim 6, characterized in that: In S1, the multimodal image acquisition module transmits the acquired image data to the computer processing system in real time to ensure the accuracy and timeliness of the image data.

8. The method for using the positioning puncture surgical robot system with reconfigurable workspace according to claim 6, characterized in that: In S2, the automatic segmentation algorithm in the three-dimensional anatomical model reconstruction module is based on deep learning technology, which is used to automatically identify and segment important structures and provide a high-precision three-dimensional model.

9. The method for using the positioning puncture surgical robot system with reconfigurable workspace according to claim 6, characterized in that: In S5, the texture feature extraction and analysis module can mark the safety of the path by calculating the minimum distance between points on the path and important structures, and detecting whether the path directly passes through important structures.

10. The method for using the positioning puncture surgical robot system with reconfigurable workspace according to claim 6, characterized in that: In S6, the path screening and optimization module comprehensively evaluates the path safety, trauma minimization, and puncture angle factors to screen out one or more optimal puncture paths.

Images