Real-time monitoring puncture robot tumor ablation method and system

Abstract

The application discloses a kind of real-time monitoring puncture robot tumor ablation method and system, method includes: obtaining tumor molecular level image, the tumor molecular level image is combined with medical image data, multi-level reconstruction is carried out, and the three-dimensional model of organ and tumor is obtained;Extract the feature point of tumor and surrounding anatomical structure, the tumor position in the three-dimensional model is accurately registered with ablation needle;Adaptive path planning algorithm is used, and the optimal insertion path of puncture needle is formulated;Sensor network is integrated on ultrasonic probe and the puncture needle, the information of needle tip and surrounding tissue is monitored, the three-dimensional model is updated heat field distribution and the real-time position of needle, and the power and time of ablation needle are adjusted;After ablation, combined examination of ultrasound and CT image is carried out, and the tumor coverage of ablation region is detected.The application can provide more accurate real-time guidance during puncture process, significantly reduce system cost, and improve the convenience of operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a real-time monitoring puncture robot tumor ablation method and system. BACKGROUND

[0002] In recent years, the progress of medical imaging technology has driven the rapid development of image-guided percutaneous puncture technology, making it a research hotspot in minimally invasive surgery and widely used in the detection and treatment of tumor or cancer patients. Traditional tumor puncture surgery and interventional surgery usually rely on preoperative images such as CT, MRI, ultrasound, etc. to understand the patient's lesion situation and plan the puncture path. During the operation, the doctor avoids important organs and blood vessel structures according to the preoperative planned path, and accurately sends the puncture needle to the target position to sample or treat the lesion site.

[0003] The three-dimensional reconstruction of preoperative images plays an important role in this process. Through three-dimensional reconstruction, doctors can more intuitively understand the internal anatomical structure of the human body, which helps to plan a more accurate surgical path. However, some problems in traditional puncture surgery, such as hand-eye discoordination, loss of direction, and difficulty in directly observing the internal structure of complex organs, bring great challenges to the operation of doctors, especially for young doctors with insufficient experience, who face a long learning period when learning how to build three-dimensional spatial perception based on two-dimensional images, affecting the efficiency and accuracy of the operation.

[0004] Traditional three-dimensional CT images can provide clear three-dimensional anatomical images before surgery, but cannot be updated and guided in real time during surgery. While two-dimensional ultrasound images have real-time imaging capabilities, they have obvious shortcomings in accurately constructing three-dimensional models, especially in terms of image clarity and noise limitations, which bring certain difficulties to surgical guidance.

[0005] In this process, surgical navigation systems have gradually become an important tool to solve this problem, combining preoperative images with intraoperative operation processes, through image registration and real-time positioning technology, to help doctors understand the spatial relationship between surgical tools and target positions in real time. Foreign surgical navigation technology started early and has developed relatively maturely, but most systems face problems such as high cost, complex operation and language barriers, affecting their popularity.

[0006] Many advanced surgical navigation systems currently rely on expensive optical positioning devices, which are not only expensive, but also susceptible to obstruction during use and require a high operating environment. In addition, the use and maintenance of imported navigation systems also face language, technical adaptability and maintenance response speed problems, which is not conducive to widespread promotion in China. Moreover, although there has been some research on surgical navigation systems in China, most of the achievements have not yet formed mature clinical applications, and the popularization rate is low, especially the lack of low-cost and easy-to-operate navigation systems that adapt to the characteristics of domestic surgical environment. SUMMARY

[0007] Therefore, the purpose of the embodiments of the present application is to provide a real-time monitoring puncture robot tumor ablation method and system, which can provide more accurate real-time guidance during puncture, significantly reduce system cost, improve operation convenience, and help solve the problems of non-real-time surgical guidance, dependence on experience, high cost and the like in traditional technologies.

[0008] In a first aspect, the embodiments of the present application provide a real-time monitoring puncture robot tumor ablation method, which comprises:

[0009] The quantum coherent imaging technology is used to obtain a tumor molecular level image, the tumor molecular level image is combined with medical image data, multi-level reconstruction is performed, and a three-dimensional model of an organ and a tumor is obtained.

[0010] Feature points of the tumor and surrounding anatomical structures are extracted, and a registration algorithm is used to perform precision registration on the tumor position in the three-dimensional model and the ablation needle.

[0011] An adaptive path planning algorithm is used to formulate an optimal insertion path of the puncture needle.

[0012] A sensor network is integrated on an ultrasonic probe and the puncture needle to monitor information of the needle tip and surrounding tissues, the three-dimensional model is updated in combination with a heat field distribution and a real-time position of the needle, and the power and time of the ablation needle are adjusted.

[0013] After ablation, ultrasonic and CT image joint inspection is performed to detect tumor coverage in the ablation area.

[0014] In combination with the first aspect, the embodiments of the present application provide a first possible implementation manner of the first aspect, wherein the quantum coherent imaging technology is used to obtain a tumor molecular level image, the tumor molecular level image is combined with medical image data, multi-level reconstruction is performed, and a three-dimensional model of an organ and a tumor is obtained, which comprises:

[0015] A quantum dot contrast agent is injected into a tumor area by using a syringe.

[0016] Quantum coherent imaging technology is adopted to scan the lesion using quantum sensors to obtain signal data reflecting the molecular characteristics of the tumor and generate a tumor molecular level image.

[0017] The lesion is scanned using computed tomography and magnetic resonance imaging technology to obtain medical image data of the anatomical structure.

[0018] A multi-modal registration algorithm based on mutual information is adopted to register the tumor molecular level image and the medical image data in the same coordinate system, so that the molecular level information and the anatomical structure information are spatially aligned to obtain fused multi-modal image data.

[0019] A visualization toolkit is used to perform three-dimensional reconstruction on the fused multi-modal image data.

[0020] A voxel-based marching cubes algorithm is adopted to mesh the surface of the tumor and the surrounding tissue to generate a three-dimensional model of the organ and the tumor.

[0021] The technical effects are that by using quantum coherent imaging technology to obtain a tumor molecular level image, quantum coherent imaging can be optimized for tumor-specific molecular signals, improving the accuracy of tumor detection; by multi-modal fusion and three-dimensional reconstruction of quantum coherent imaging data and anatomical image data such as CT and MRI, and fine meshing of the surface of the tumor and its surrounding tissue, the differences in different image data sources are effectively handled, a clear and accurate three-dimensional model is generated, the quality and detail performance of three-dimensional reconstruction are improved, and the internal molecular characteristic distribution can be displayed to provide more comprehensive information support for doctors in surgical planning and decision-making.

[0022] In combination with the first aspect, the embodiments of the present application provide a second possible implementation manner of the first aspect, wherein the extracting feature points of the tumor and the surrounding anatomical structure and performing precision registration of the tumor position in the three-dimensional model and the ablation needle using a registration algorithm comprise:

[0023] A feature-based detection algorithm is adopted to extract feature points of the tumor and the surrounding anatomical structure in the three-dimensional model.

[0024] Random sample consensus algorithm is used to eliminate outliers to obtain an effective feature point set.

[0025] A group of initial feature points is selected from the effective feature point set, and a nearest neighbor algorithm is used to quickly register the feature points in the effective feature point set with the initial position feature points of the ablation needle to obtain a coarse registration three-dimensional model.

[0026] An iterative closest point algorithm based on mutual information is adopted to perform fine registration on the result of fast registration to obtain a fine registration three-dimensional model.

[0027] The technical effect is that the feature-based detection algorithm, the random sample consensus, the nearest neighbor algorithm and the mutual information-based ICP algorithm are combined to realize high-precision registration of the tumor position and the ablation needle, and the complexity of manual adjustment of the ablation needle position by the doctor during the operation is reduced. Due to the accurate registration of the ablation needle and the tumor position, the system can adjust the ablation power and time according to the real-time feedback, ensure that the ablation range covers the entire tumor tissue, and avoid damage to the surrounding healthy tissue caused by excessive ablation.

[0028] In combination with the first aspect, the embodiments of the first aspect provide a third possible implementation manner of the first aspect, wherein the adaptive path planning algorithm is used to formulate the optimal insertion path of the puncture needle.

[0029] In the three-dimensional model, the initial position P0 of the puncture needle, the tumor ablation area A T , and the non-puncture area A B are marked.

[0030] The initial position P0 of the puncture needle is taken as a root node V0 of a path tree, and a rapid expansion random tree algorithm is used to initialize the path tree T as an empty tree in the three-dimensional model.

[0031] The maximum iteration number and the random sampling range are set, and a random node is generated in the three-dimensional model.

[0032] For each generated random node, all nodes in the path tree T are traversed to find the nearest path tree node to the random node.

[0033] A new random node is generated by expanding a fixed step length from the nearest path tree node towards the random node, the path tree T is repeatedly traversed, and the corresponding path tree node is generated.

[0034] Before adding the new random node to the path tree T, a path interpolation point q t =q near +t·(q new -q near )(0≤t≤1) is calculated, q new is the new random node, q near is the corresponding path tree node, and it is checked whether each path interpolation point q t is located in the non-puncture area A B .

[0035] If the corresponding path does not collide, the new random node q new is added to the path tree T.

[0036] repeating the generation of the path tree nodes until the maximum number of iterations is reached, or the distance between a node in the path tree T and the tumor ablation region A T is less than a preset threshold.

[0037] When the path tree T extends to the tumor ablation region A T , backtracking from the new node of the extension to the root node V0 obtains an optimal insertion path of the puncture needle.

[0038] The technical effect is that a fast expanding random tree algorithm is used to construct a path tree from the initial position of the puncture needle, and a path is gradually generated and expanded, so that a feasible puncture path can be efficiently found in a complex three-dimensional anatomical structure. By selecting the nearest node in the path tree for expansion at each expansion, the generation of the path is more targeted, thereby shortening the time to reach the target. By calculating the path interpolation point and checking whether it falls into an impassable region, collision between the puncture path and surrounding important anatomical structures (such as blood vessels, nerves, etc.) can be effectively detected and avoided. By obtaining the optimal insertion path of the puncture needle, the puncture process can accurately reach the best position of the tumor region, thereby ensuring that the ablation needle can reach the predetermined ablation position in the shortest path and maximize the coverage of the tumor tissue.

[0039] In combination with the first aspect, the fourth possible implementation of the first aspect is provided, wherein the sensor network integrated on the ultrasound probe and the puncture needle monitors information of the needle tip and the surrounding tissue, and the power and time of the ablation needle are adjusted by combining the three-dimensional model to update the heat field distribution and the real-time position of the needle, which comprises:

[0040] The sensor network integrated on the ultrasound probe and the puncture needle includes temperature sensors, pressure sensors, position sensors, and ultrasonic sensors.

[0041] The real-time image data obtained by the ultrasonic sensor is registered with the three-dimensional model to update the accurate position of the puncture needle in the three-dimensional space, the puncture angle, and the state of the surrounding tissue.

[0042] A heat conduction model is established by using the data obtained by the temperature sensor wherein p is the tissue density, Q is the heat source power of the ablation needle, k is the thermal conductivity of the tissue, and c is the specific heat capacity of the tissue, and the real-time temperature distribution T(x, y, z, t) in the tumor ablation region is solved.

[0043] Based on the real-time temperature distribution T(x, y, z, t), a temperature threshold is set, if the temperature of the current ablation region is lower than the temperature threshold, the output power of the ablation needle is increased, if the temperature is higher than the temperature threshold, the output power of the ablation needle is reduced.

[0044] The technical effects are that real-time image data acquired by the ultrasonic sensor is registered with the three-dimensional model, the position and angle of the puncture needle in the three-dimensional space can be updated in real time, real-time accurate positioning and feedback are realized, data acquired by the temperature sensor can be combined with parameters such as tissue density, heat source power of the ablation needle, thermal conductivity and specific heat capacity to optimize heat field distribution, accurate ablation is realized, based on real-time temperature distribution, dynamic power adjustment can be performed to avoid excessive damage to the tissue, through automatic power adjustment and real-time monitoring of heat field distribution, accurate heating of the tumor area is realized, ablation efficiency is effectively improved, and ablation time is shortened.

[0045] With reference to the first aspect, the embodiments of the present application provide a fifth possible implementation manner of the first aspect.

[0046] In the ablation process, the position of the current ablation area acquired in real time is compared with the boundary of the tumor ablation area through the sensor network.

[0047] If the deviation distance of the current ablation area from the boundary of the tumor ablation area exceeds a set threshold value, it is judged that the current ablation area deviates from the tumor ablation area, and the current position of the puncture needle is taken as a root node of a path tree, and the puncture path is recalculated.

[0048] The technical effects are that when the deviation distance exceeds the set threshold value, the system automatically judges that the ablation area has deviated from the target area, and takes corresponding adjustment measures to compensate for the deviation faster and more accurately, and avoid unnecessary damage to healthy tissues due to deviation of the ablation needle.

[0049] With reference to the first aspect, the embodiments of the present application provide a sixth possible implementation manner of the first aspect.

[0050] For the puncture needle and the ablation needle, a multi-needle cooperative ablation technology is adopted.

[0051] According to the three-dimensional model and the data collected by the sensor network, a three-dimensional finite element simulation model is established.

[0052] Using finite element analysis, heat conduction in the ablation process is simulated in the three-dimensional finite element simulation model to predict the heat field distribution of the ablation needle at different insertion positions and angles.

[0053] By setting an objective function, the optimal insertion position and the optimal insertion angle of the multi-needle are calculated by using an optimization algorithm, aiming at maximizing tumor coverage and minimizing healthy tissue damage.

[0054] Based on the optimal insertion position and the optimal insertion angle, an adaptive path planning algorithm is used to formulate an optimal insertion path for each of the puncture needles, and to avoid non-puncturable areas.

[0055] The real-time positions of the needles and the surrounding tissue states are monitored by the integrated sensor network, and if it is monitored that the ablation area deviates from the expected target, the path of the corresponding needle is recalculated so that the ablation area covers the tumor.

[0056] The technical effect is that the multi-needle cooperative ablation technology can simultaneously perform thermal ablation treatment at multiple positions, effectively improving the efficiency of ablation, and at the same time, the insertion position and angle of each needle are flexibly adjusted during the operation, thereby better adapting to the shape and size of the tumor and improving the adaptability to irregular tumors.

[0057] In combination with the first aspect, embodiments of the first aspect provide a seventh possible implementation, wherein after the ablation is completed, a combined examination of ultrasound and CT images is performed to detect the tumor coverage of the ablation area, including:

[0058] Real-time ultrasound images and CT images of the area after the ablation are obtained, and the ultrasound images and the CT images are image-registered to obtain a fusion image.

[0059] A segmentation method is used to extract the ablation area and the un-ablated tumor tissue from the fusion image.

[0060] The three-dimensional model reconstructed before ablation is spatially superimposed on the extracted ablation area to calculate the coverage rate of the ablation area on the tumor.

[0061] If the coverage rate reaches a set threshold, the ablation operation is determined to be effective; otherwise, the residual tumor position is marked and the treatment result is fed back.

[0062] The technical effect is that through the combined examination of ultrasound and CT images, image registration, segmentation, spatial superposition, and coverage rate calculation are used to realize accurate evaluation of the ablation area and un-ablated tumor tissue, and to provide an automatic postoperative effect judgment and feedback mechanism, thereby improving the detection accuracy of the ablation area and realizing accurate identification of tumor residues to provide accurate targeting information for subsequent supplementary ablation or other treatment methods, and to help doctors develop more targeted treatment plans.

[0063] In combination with the first aspect, embodiments of the first aspect provide an eighth possible implementation, wherein after the ablation is completed, the method further includes:

[0064] Thermal conduction data of the area after the ablation is obtained, and the thermal conduction data is fused with the three-dimensional model to establish a tumor thermal field distribution model.

[0065] Generate a personalized postoperative care and review plan in combination with the patient's tumor biological parameters.

[0066] The technical effect is that by acquiring heat conduction data after ablation and fusing with a three-dimensional model, a personalized postoperative care and review plan is generated in combination with the patient's biological parameters, precise evaluation of ablation effect and personalized postoperative management are achieved, postoperative rehabilitation quality is improved, review plan is optimized, the risk of complications is reduced, and data-driven treatment effect optimization is promoted, providing more scientific and effective postoperative management services for patients.

[0067] In a second aspect, the embodiment of the present application also provides a real-time monitoring puncture robot tumor ablation system, which comprises:

[0068] The lesion modeling module is used to acquire tumor molecular level images by quantum coherent imaging technology, combine the tumor molecular level images with medical image data, perform multi-level reconstruction, and obtain a three-dimensional model of an organ and a tumor.

[0069] The feature point matching module is used to extract feature points of a tumor and surrounding anatomical structures, and perform precision registration of the tumor position in the three-dimensional model and the ablation needle by using a registration algorithm.

[0070] The path design module is used to develop an optimal insertion path of the puncture needle by using an adaptive path planning algorithm.

[0071] The sensor module is used to integrate a sensor network on an ultrasonic probe and the puncture needle, monitor information of a needle tip and surrounding tissues, update a heat field distribution and a real-time position of the needle in combination with the three-dimensional model, and adjust the power and time of the ablation needle.

[0072] The joint inspection module is used to perform joint inspection of ultrasonic and CT images after ablation is completed, and detect tumor coverage in the ablation area.

[0073] The embodiment of the present application has the following beneficial effects:

[0074] The present application acquires tumor molecular level images by quantum coherent imaging technology, combines traditional medical images such as CT and MRI, performs multi-modal registration and three-dimensional reconstruction, and accurately displays the morphology of the tumor and the surrounding anatomical structure.

[0075] The present application can automatically generate an optimal puncture path by using an adaptive path planning algorithm such as a rapidly-exploring random tree, avoids collision risks in non-punctureable areas, and improves the safety of the operation.

[0076] The present application can monitor the state of the puncture needle and the tissue in real time through a sensor network, integrate temperature, pressure and position sensors on the puncture needle and the ultrasound probe, monitor the position of the needle tip and the heat field distribution in real time, and dynamically adjust the power and time of the ablation needle according to the information, so as to more accurately control the ablation range and effect.

[0077] The present application can optimize the multi-needle insertion position and angle, improve the tumor coverage rate and reduce the damage to healthy tissues through multi-needle cooperative ablation and finite element simulation analysis technology.

[0078] After the ablation is completed, the present application can accurately detect the tumor coverage through the combined examination of ultrasound and CT images, and generate a personalized postoperative care and review scheme based on the heat conduction data and the biological parameters of the patient. BRIEF DESCRIPTION OF DRAWINGS

[0079] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0080] Fig. 1 Flow chart of the puncture robot tumor ablation method of the present application for real-time monitoring;

[0081] Fig. 2 Technical route schematic diagram of the puncture robot tumor ablation method of the present application for real-time monitoring. DETAILED DESCRIPTION

[0082] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0083] Please refer to Figs. 1-2The first embodiment of the present application provides a real-time monitoring puncture robot tumor ablation method, comprising: acquiring tumor molecular level images by quantum coherent imaging technology, combining the tumor molecular level images with medical image data, performing multi-level reconstruction to obtain a three-dimensional model of an organ and a tumor; extracting feature points of the tumor and surrounding anatomical structures, and performing precision registration of the tumor position in the three-dimensional model and an ablation needle by using a registration algorithm; using an adaptive path planning algorithm to formulate an optimal insertion path of the puncture needle; integrating a sensor network on an ultrasound probe and the puncture needle to monitor information of a needle tip and surrounding tissues, combining the three-dimensional model to update a thermal field distribution and a real-time position of the needle, and adjusting power and time of the ablation needle; after ablation is completed, performing joint examination of ultrasound and CT images to detect tumor coverage in an ablation area.

[0084] In combination with the first aspect, the embodiments of the present application provide a first possible implementation manner of the first aspect, wherein the acquiring tumor molecular level images by quantum coherent imaging technology, combining the tumor molecular level images with medical image data, and performing multi-level reconstruction to obtain a three-dimensional model of an organ and a tumor comprises: using a syringe to inject quantum dot contrast agents into a tumor area to improve contrast of the tumor and surrounding normal tissues; using quantum coherent imaging technology, using a quantum sensor to scan a lesion to acquire signal data reflecting tumor molecular characteristics, and generating tumor molecular level images; the quantum sensor uses quantum dots or superconducting quantum interference devices to detect small magnetic field changes inside the tumor; using computed tomography (CT) and magnetic resonance imaging (MRI) technology to scan the lesion to obtain medical image data of anatomical structures; CT images are mainly used to accurately display the relative position relationship between the tumor position and the organ, and MRI images are used to provide high-contrast information of soft tissues; using a multi-modal registration algorithm based on mutual information, the tumor molecular level images and the medical image data are registered in the same coordinate system, so that the molecular level information and the anatomical structure information are spatially aligned to obtain fused multi-modal image data; using a visualization toolkit to perform three-dimensional reconstruction on the fused multi-modal image data; using a voxel-based marching cubes algorithm to grid the surface of the tumor and surrounding tissues to generate a three-dimensional model of the organ and the tumor.

[0085] In combination with the molecular information provided by quantum coherent imaging, high metabolic areas, low metabolic areas, and necrotic areas inside the tumor can be distinguished in the three-dimensional model.

[0086] Specifically, the step of using the visualization toolkit to perform three-dimensional reconstruction on the fused multi-modal image data includes: importing the fused multi-modal image data (including CT, MRI and quantum coherence imaging results) into the visualization toolkit VTK, storing in DICOM format, reading and importing using vtkDICOMImageReader, performing image enhancement, noise filtering and other preprocessing to ensure reconstruction quality; using vtkImageData to store the preprocessed multi-modal image data, setting the size of the voxel according to the required spatial resolution wherein v x,y,z represents the intensity value of the voxel in three-dimensional space, which is used to reflect the different densities of the tumor and the surrounding tissue; the volume data is mapped into three-dimensional space using vtkVolume and vtkVolumeMapper, the density and transparency are adjusted through different transfer functions, and the boundaries of the tumor and the surrounding tissue are highlighted, in specific embodiments, the transfer function presents the high-density area (such as bone) and the low-density area (such as soft tissue) in different colors, thereby improving the readability of the image.

[0087] Specifically, the step of using the voxel-based marching cube algorithm to mesh the surface of the tumor and the surrounding tissue to generate a three-dimensional model of the organ and the tumor includes: setting an isosurface threshold value α for distinguishing the tumor from the normal tissue, wherein f(x,y,z) represents the binarization result of the voxel in three-dimensional space, according to the setting of α, the boundary of the tumor tissue or the organ structure is extracted; each cube unit of the volume data is traversed, the cube unit is composed of eight adjacent voxel points, and the binarization result of the voxel points is matched with the corresponding triangular mesh in the lookup table, wherein v i represents the binarization result of the eight vertices of the unit cube, and Index is used to find the appropriate triangular configuration in the lookup table; each unit cube is converted into a plurality of triangles according to the triangular configuration in the lookup table, and a complete three-dimensional mesh of the surface of the tumor and the organ is generated by splicing a plurality of the triangles.

[0088] Further including using three-dimensional slice reconstruction software to optimize and adjust the surface smoothness and detail level of the three-dimensional model to ensure the accuracy of the tumor and its surrounding tissue structure, and the specific steps include: smoothing the triangular mesh to eliminate sharp edges and noise on the surface, adjusting the position of each vertex to approach the average value of the neighboring vertices, wherein v′ i is the smoothed vertex position, λ is the smoothing coefficient, and N(i) is the vertex v iThe neighborhood of the tumor; on the basis of the smoothing processing, the grid is optimized using a Decimation algorithm to reduce the number of triangles to improve the processing efficiency, and key structural details are reserved; the surface reconstruction function of 3DSlicer is used to generate a three-dimensional model with better visualization effect, so that the detail level is reserved when the tumor and the surrounding tissue are displayed, and more accurate surgical reference is provided.

[0089] The technical effect is that: by using quantum coherent imaging technology to obtain tumor molecular level images, quantum coherent imaging can be optimized for specific molecular signals of tumors, improving the accuracy of tumor detection; by multi-modal fusion and three-dimensional reconstruction of quantum coherent imaging data and CT and MRI anatomical image data, the surface of the tumor and its surrounding tissue is finely meshed, effectively dealing with the differences in different image data sources, generating a clear and accurate three-dimensional model, improving the quality and detail performance of three-dimensional reconstruction, and showing the internal molecular feature distribution, helping doctors have more comprehensive information support in surgical planning and decision-making.

[0090] In combination with the first aspect, the embodiments of the present application provide a second possible implementation manner of the first aspect, wherein the extracting feature points of the tumor and the surrounding anatomical structure, and performing precision registration of the tumor position in the three-dimensional model and the ablation needle by using a registration algorithm, comprises: extracting feature points of the tumor and the surrounding anatomical structure in the three-dimensional model by using a feature-based detection algorithm, specifically, using a SIFT (Scale-Invariant Feature Transform) or SURF (Speeded-Up Robust Features) feature detection algorithm to extract feature points of the tumor and the surrounding anatomical structure from the three-dimensional model, each feature point generates a descriptor, which is a local image feature vector, used for subsequent matching and registration; using a random sample consensus algorithm to eliminate outliers and select an effective feature point set, specifically, the random sample consensus algorithm iteratively selects feature point pairs, fits a model, and evaluates the inliers and outliers of the model, sets a threshold, selects an effective feature point set, eliminates outliers that do not meet the model, and outputs the effective feature point set P valid ={p1,p2,...,p n};select a group of initial feature points from the effective feature point set, use the nearest neighbor algorithm to quickly register the feature points in the effective feature point set with the initial position feature points of the ablation needle, and obtain a coarse registration three-dimensional model, specifically, select a group of initial feature points from the effective feature point set P init ={p i1 ,p i2 ,...,p ik}, calculate the distance d(p i ,p j), using a nearest neighbor algorithm to match the feature points in the effective feature point set with the initial position feature points of the ablation needle, for each feature point p i In the effective feature point set P valid , find the nearest neighbor point in the ablation needle feature point set Record the matching results to form a coarse registration three-dimensional model M coarse ; the result of the fast registration based on mutual information is used for fine registration to obtain a fine registration three-dimensional model, specifically, the coarse registration three-dimensional model M coarse is taken as the initial input, and the registration accuracy is gradually improved through an iterative optimization method, and in each iteration, the mutual information of the two groups of points is calculated The model is updated according to the change amount of the mutual information, the transformation matrix T is optimized, and the fine registration three-dimensional model M fine is obtained until convergence.

[0091] By overlapping and comparing the fine registration three-dimensional model with the initial three-dimensional model, the overlap degree is calculated, and the registration accuracy is quantitatively evaluated, specifically, the voxel overlap rate index is used for evaluation,

[0092] The technical effects are that: by combining the feature-based detection algorithm, the random sample consensus, the nearest neighbor algorithm and the ICP algorithm based on mutual information, high-precision registration of the tumor position and the ablation needle is realized, and the complexity of manual adjustment of the ablation needle position by the doctor during the operation is reduced. Due to the accurate registration of the ablation needle and the tumor position, the system can adjust the ablation power and time according to real-time feedback, ensure that the ablation range covers the entire tumor tissue, and at the same time avoid excessive ablation to cause damage to the surrounding healthy tissue.

[0093] In combination with the first aspect, the embodiments of the present application provide a third possible implementation manner of the first aspect, wherein the adaptive path planning algorithm is used to formulate the optimal insertion path of the puncture needle, including: in the three-dimensional model, marking the initial position P0 of the puncture needle, the tumor ablation area A T , and the non-puncture area A B ; taking the initial position P0 of the puncture needle as the root node V0 of the path tree, initializing the path tree T as an empty tree in the three-dimensional model; setting the maximum iteration number and the random sampling range, and generating random nodes in the three-dimensional model in a loop; for each generated random node, traversing all nodes in the path tree T, finding the nearest path tree node to the random node; from the nearest path tree node to the random node, a fixed step is expanded to generate a new random node, and the path tree T is repeatedly traversed and the corresponding path tree node is generated; before adding the new random node to the path tree T, the path interpolation point q t=q near +t·(q new -q near (0≤t≤1), q new For the new random node, q near For the corresponding path tree node, use linear interpolation to check whether the path from the nearest path tree node to the new random node is adjacent to the impenetrable region A. B Cross-check each of the path interpolation points q t Is it located in the non-puncture area A? B If a collision occurs, the corresponding path is considered a collision; if the corresponding path does not collide, the new random node q is... new Add to the path tree T; repeat the generation of the path tree nodes until the maximum number of iterations is reached, or the nodes in the path tree T are associated with the tumor ablation region A. T The distance is less than a preset threshold; when the path tree T extends to the tumor ablation area A T By tracing back from the expanded new node to the root node V0, the optimal insertion path of the puncture needle is obtained.

[0094] The technical advantages are as follows: A fast expanding random tree algorithm is used to construct a path tree starting from the initial position of the puncture needle, progressively generating and expanding the path to efficiently find a feasible puncture path in complex three-dimensional anatomical structures. During each expansion, the nearest node in the path tree is selected for expansion, making path generation more targeted and shortening the time to reach the target. By calculating path interpolation points and checking for non-puncture areas, collisions between the puncture path and surrounding important anatomical structures (such as blood vessels and nerves) are effectively detected and avoided. By obtaining the optimal insertion path of the puncture needle, the puncture process can accurately reach the optimal position in the tumor area, ensuring that the ablation needle reaches the predetermined ablation location within the shortest path and maximizes tumor tissue coverage.

[0095] In conjunction with the first aspect, this invention provides a fourth possible implementation of the first aspect, wherein integrating a sensor network on the ultrasonic probe and the puncture needle to monitor information about the needle tip and surrounding tissue, updating the thermal field distribution and the real-time position of the needle in conjunction with the three-dimensional model, and adjusting the power and time of the ablation needle includes: integrating a sensor network on the ultrasonic probe and the puncture needle, including a temperature sensor, a pressure sensor, a position sensor, and an ultrasonic sensor; registering the real-time image data acquired by the ultrasonic sensor with the three-dimensional model to update the precise position of the puncture needle in three-dimensional space, the puncture angle, and the state of the surrounding tissue; and establishing a heat conduction model using the data acquired by the temperature sensor. Wherein, p is the tissue density, Q is the heat source power of the ablation needle, k is the thermal conductivity of the tissue, c is the specific heat capacity of the tissue, and the real-time temperature distribution T(x, y, z, t) in the tumor ablation area is solved; based on the real-time temperature distribution T(x, y, z, t), a temperature threshold is set, if the temperature of the current ablation area is lower than the temperature threshold, the output power of the ablation needle is increased, if it is higher than the temperature threshold, the output power of the ablation needle is reduced, P(t+1)=P(t)+K p (T threshold -T current ), wherein K p is the power adjustment gain coefficient, T threshold is the temperature of the current ablation area, T current is the real-time measured tissue temperature, and P is the output power of the ablation needle.

[0096] The technical effects are that: the real-time image data acquired by the ultrasonic sensor is registered with the three-dimensional model, the position and angle of the puncture needle in the three-dimensional space can be updated in real time, real-time accurate positioning and feedback are realized; the data acquired by the temperature sensor can be combined with parameters such as tissue density, heat source power of the ablation needle, thermal conductivity and specific heat capacity to optimize the heat field distribution, realize accurate ablation, based on the real-time temperature distribution, dynamic power adjustment can be carried out to avoid excessive damage to the tissue, through automatic power adjustment and real-time monitoring of the heat field distribution, accurate heating of the tumor area is realized, the ablation efficiency is effectively improved, and the ablation time is shortened.

[0097] In combination with the first aspect, the embodiments of the first aspect provide a fifth possible implementation manner of the first aspect, and further comprise: in the ablation process, comparing the position of the current ablation area acquired in real time with the boundary of the tumor ablation area through the sensor network; if the deviation distance between the current ablation area and the boundary of the tumor ablation area exceeds a set threshold, it is judged that the current ablation area deviates from the tumor ablation area, and the current position of the puncture needle is taken as a root node of a path tree, and the puncture path is recalculated.

[0098] The technical effects are that: when the deviation distance exceeds the set threshold, the system can automatically judge that the ablation area has deviated from the target area, and take corresponding adjustment measures to compensate the deviation faster and more accurately, and avoid unnecessary damage to healthy tissues caused by deviation of the ablation needle.

[0099] With reference to the first aspect, the embodiments of the present application provide a seventh possible implementation manner of the first aspect, wherein after the ablation is completed, joint examination of ultrasound and CT images is performed to detect the tumor coverage of the ablation region, including: acquiring real-time ultrasound images and CT images of the region after the ablation is completed, performing image registration on the ultrasound images and the CT images to obtain fused images; using a segmentation method to extract the ablation region and the non-ablated tumor tissue from the fused images; performing spatial superposition of the three-dimensional model reconstructed before the ablation on the ablation region to calculate the coverage rate of the ablation region on the tumor; if the coverage rate reaches a set threshold, it is determined that the ablation operation is effective; otherwise, the residual tumor position is marked, and the treatment result is fed back.

[0100] The technical effects are that: the multi-needle cooperative ablation technology can simultaneously perform thermal ablation processing at multiple positions, effectively improving the ablation efficiency, and at the same time, the insertion position and angle of each needle can be flexibly adjusted during the operation, thereby better adapting to the shape and size of the tumor and improving the adaptability to irregular tumors.

[0101] With reference to the first aspect, the embodiments of the present application provide a seventh possible implementation manner of the first aspect, wherein after the ablation is completed, joint examination of ultrasound and CT images is performed to detect the tumor coverage of the ablation region, including: acquiring real-time ultrasound images and CT images of the region after the ablation is completed, performing image registration on the ultrasound images and the CT images to obtain fused images; using a segmentation method to extract the ablation region and the non-ablated tumor tissue from the fused images; performing spatial superposition of the three-dimensional model reconstructed before the ablation on the ablation region to calculate the coverage rate of the ablation region on the tumor; if the coverage rate reaches a set threshold, it is determined that the ablation operation is effective; otherwise, the residual tumor position is marked, and the treatment result is fed back.

[0102] The technical effects are that: through the joint examination of ultrasound and CT images, the precise evaluation of the ablation region and the non-ablated tumor tissue is realized by means of image registration, segmentation, spatial superposition and coverage rate calculation, and an automatic postoperative effect judgment and feedback mechanism is provided, the detection accuracy of the ablation region is improved, the precise identification of the residual tumor is realized, precise targeting information is provided for subsequent supplementary ablation or other treatment methods, and a more targeted treatment plan is helped to be formulated by the doctor.

[0103] With reference to the first aspect, the embodiments of the present application provide an eighth possible implementation manner of the first aspect, and the method further comprises: after the ablation is completed, acquiring heat conduction data of a region after the ablation is completed, fusing the heat conduction data with the three-dimensional model, and establishing a tumor heat field distribution model, specifically, combining the heat field distribution recorded during the ablation, the power adjustment curve, the real-time position data of the puncture needle, the heat diffusion and tissue reaction data, and the tumor characteristics (such as size, shape, tissue density, etc.) of the patient to establish a personalized tumor heat field distribution model, which reflects the heat conduction and attenuation of heat in the tumor and the surrounding tissue during the ablation, and is used to simulate the thermal effect and recovery of the tissue after the ablation; combining the biological parameters of the tumor of the patient, a personalized postoperative care and review scheme is generated, specifically, the biological parameters of the patient, such as the pathological type of the tumor, gene mutation, blood index, immune response, etc., are acquired, the recovery speed of the tissue in the ablation region, the possible residual tumor region, and the reaction degree of the patient's body to the ablation are analyzed, the recovery time of the tissue after the ablation, including the repair speed and range of the heat-damaged tissue, is predicted, the heat effect and tissue reaction after the ablation are combined to evaluate whether auxiliary treatment, such as anti-inflammatory treatment, anti-tumor drugs, etc., is needed, specific postoperative care suggestions are provided according to the recovery of the patient, the residual risk and tissue recovery condition predicted based on the heat field model are used to determine an appropriate postoperative review time point, and the imaging examination of the ablation region is performed, the morphological change of the tissue in the ablation region is detected, and whether new tumor cells or residual lesions exist is observed.

[0104] The technical effects are that the ablation effect is accurately evaluated and personalized postoperative management is realized by acquiring the heat conduction data after the ablation and fusing the heat conduction data with the three-dimensional model, and a personalized postoperative care and review scheme is generated by combining the biological parameters of the patient, the rehabilitation quality after the operation is improved, the review scheme is optimized, the risk of complications is reduced, data-driven optimization of treatment effect is promoted, and more scientific and effective postoperative management services are provided for the patient.

[0105] The second embodiment of the application provides a real-time monitoring puncture robot tumor ablation system, comprising: a lesion modeling module, which is used for acquiring tumor molecular level images by using quantum coherent imaging technology, combining the tumor molecular level images with medical image data, performing multi-level reconstruction, and obtaining a three-dimensional model of an organ and a tumor; a feature point matching module, which is used for extracting feature points of the tumor and surrounding anatomical structures, and performing precision registration on the tumor position in the three-dimensional model and an ablation needle by using a registration algorithm; a path design module, which is used for formulating an optimal insertion path of the puncture needle by using an adaptive path planning algorithm; a sensor module, which is used for integrating a sensor network on an ultrasonic probe and the puncture needle, monitoring information of a needle tip and surrounding tissues, updating a thermal field distribution and a real-time position of the needle in combination with the three-dimensional model, and adjusting power and time of the ablation needle; and a joint inspection module, which is used for performing joint inspection of ultrasonic and CT images after ablation is completed, and detecting tumor coverage of an ablation area.

[0106] The computer program product of the real-time monitoring puncture robot tumor ablation method and device provided by the embodiment of the application comprises a computer readable storage medium storing program codes, and the instructions included in the program codes can be used to execute the method in the foregoing method embodiment.

[0107] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, the computer program can execute the real-time monitoring puncture robot tumor ablation method described above, so that more accurate real-time guidance can be provided in the puncture process, the system cost is significantly reduced, the operation convenience is improved, and the problems such as non-real-time surgery guidance, dependence on experience, and high cost in the traditional technology can be solved.

[0108] If the functions are realized in the form of software function units and sold or used as independent products, the functions can be stored in a nonvolatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the application or parts of the technical solutions that essentially contribute to the prior art or parts of the technical solutions can be embodied in the form of software products. The computer software product is stored in a storage medium and includes instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0109] Finally, it should be noted that the above-described embodiments are merely specific embodiments of the present application, which are used to illustrate the technical solutions of the present application, but not to limit the same. The protection scope of the present application is not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that any person skilled in the art can still modify or easily think of changes to the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some of the technical features, within the technical scope disclosed by the present application. The modifications, changes or replacements do not cause the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A real-time monitored, puncture robot tumor ablation system, characterized in that, The method comprises the following steps: a lesion modeling module is used to acquire tumor molecular level images by quantum coherence imaging technology, combine the tumor molecular level images with medical image data, perform multi-level reconstruction, and obtain three-dimensional models of organs and tumors; a feature point matching module is used to extract feature points of tumors and surrounding anatomical structures, and perform precision registration of tumor positions in the three-dimensional models and ablation needles by using a registration algorithm; A path design module is configured to formulate an optimal insertion path of the puncture needle by using an adaptive path planning algorithm, and specifically includes: marking an initial position of the puncture needle in the three-dimensional model , a tumor ablation region , and an impassable region ; taking the initial position of the puncture needle as a root node of a path tree , initializing the path tree in the three-dimensional model by using a rapidly-exploring random tree algorithm to generate a path tree ; setting a maximum iteration number and a random sampling range, and cyclically generating random nodes in the three-dimensional model; for each generated random node, traversing all nodes in the path tree to find a path tree node closest to the random node; expanding a fixed step length from the closest path tree node toward the random node to generate a new random node, repeating the traversal of the path tree and generating a corresponding path tree node; before adding the new random node to the path tree , calculating path interpolation points for the new random node and the corresponding path tree node, checking whether each path interpolation point is located in the impassable region , if so, regarding the corresponding path as a collision; if the corresponding path has no collision, adding the new random node to the path tree ; repeating the generation of the path tree node until the maximum iteration number is reached, or the distance between a node in the path tree and the tumor ablation region is less than a preset threshold; when the path tree extends to the tumor ablation region , backtracking from the extended new node to the root node to obtain the optimal insertion path of the puncture needle. A sensor module is used to integrate a sensor network on the ultrasound probe and the puncture needle, monitor information of the needle tip and the surrounding tissue, update the heat field distribution and the real-time position of the needle in combination with the three-dimensional model, and adjust the power and time of the ablation needle, specifically including: integrating a sensor network on the ultrasound probe and the puncture needle, including temperature sensors, pressure sensors, position sensors and ultrasonic sensors; registering real-time image data acquired by the ultrasonic sensor with the three-dimensional model, updating the accurate position of the puncture needle in three-dimensional space, the puncture angle and the state of the surrounding tissue; using the data acquired by the temperature sensor to establish a heat conduction model wherein, is the tissue density, is the heat source power of the ablation needle, is the thermal conductivity of the tissue, is the specific heat capacity of the tissue, and the real-time temperature distribution in the tumor ablation area is solved ; based on the real-time temperature distribution , setting a temperature threshold, if the temperature of the current ablation area is lower than the temperature threshold, increasing the output power of the ablation needle, if higher than the temperature threshold, reducing the output power of the ablation needle; further comprising, during the ablation process, comparing the real-time acquired position of the current ablation area with the boundary of the tumor ablation area through the sensor network, if the deviation distance of the current ablation area from the boundary of the tumor ablation area exceeds a set threshold, judging that the current ablation area deviates from the tumor ablation area, taking the current position of the puncture needle as the root node of the path tree, and recalculating the puncture path; after the accuracy registration, further comprising, for the puncture needle and the ablation needle, using a multi-needle collaborative ablation technology, establishing a three-dimensional finite element simulation model according to the three-dimensional model and the data collected by the sensor network, using finite element analysis to simulate heat conduction in the three-dimensional finite element simulation model during ablation, predicting the heat field distribution of the ablation needle under different insertion positions and angles, setting an objective function to maximize the tumor coverage rate and minimize the healthy tissue damage as the target, using an optimization algorithm to calculate the best insertion position and the best insertion angle of the multi-needle, and using an adaptive path planning algorithm based on the best insertion position and the best insertion angle to develop the optimal insertion path of each puncture needle and avoid the non-puncturable area, monitoring the real-time position of each needle and the state of the surrounding tissue through the integrated sensor network, and recalculating the path of the corresponding needle if the ablation area deviates from the expected target, so that the ablation area covers the tumor; a joint inspection module is used to perform joint inspection of ultrasound and CT images after ablation, detect tumor coverage of the ablation area, and specifically comprises the following steps: acquiring real-time ultrasound images and CT images of the ablation area, performing image registration on the ultrasound images and the CT images to obtain fusion images, extracting the ablation area and un-ablated tumor tissues from the fusion images by using a segmentation method, superimposing the three-dimensional model reconstructed before ablation and the extracted ablation area in space, calculating the coverage rate of the ablation area on the tumor, determining that the ablation operation is effective if the coverage rate reaches a set threshold, otherwise marking the position of residual tumor and feeding back the treatment result.

Images