Multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device and method for fresh cadaver heads

Abstract

This application discloses a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device and method for a fresh cadaver head. The method is applied to the control module of the device, which also includes an endoscope module and an electromagnetic tracking module. The method includes: obtaining multimodal image information corresponding to the fresh cadaver head to be measured; generating a three-dimensional model of the fresh cadaver head based on the multimodal image information; obtaining a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of the nasal cavity, sinuses, blood vessels, and nerves; when the endoscope module enters the nasal cavity of the fresh cadaver head, obtaining magnetic field information sent by the electromagnetic tracking module; calculating the spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information; generating navigation information corresponding to the endoscope module based on the spatial position and the target position, the navigation information including direction information and distance information; and moving the endoscope module to the target position based on the navigation information.

Description

Technical Field

[0001] The present application relates to the field of electromagnetic navigation technology, and in particular to a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device and method for a fresh cadaver head. Background Art

[0002] In modern medicine, particularly in otolaryngology and neurosurgery, endoscopic endoscopic approach (EEA) has become an important minimally invasive surgical technique. EEA, which accesses the skull base through the nasal cavity, can be used to treat a variety of conditions, such as pituitary tumors and meningiomas. However, due to the complex anatomy of the nasal cavity and paranasal sinuses and their significant individual variability, highly precise navigation and positioning systems are required to ensure safety and effectiveness during surgery.

[0003] Traditional navigation systems rely primarily on preoperative imaging data (such as CT and MRI) for positioning, but these systems have the following limitations:

[0004] 1. Low image registration accuracy: The registration error between image data of different modalities is large, affecting navigation accuracy.

[0005] 2. Insufficient real-time feedback: Traditional systems lack real-time dynamic update capabilities and cannot adjust according to changes during the surgical process.

[0006] 3. Complex operation: Key anatomical structures need to be manually marked, which increases the workload of doctors.

[0007] 4. Security issues: In a complex environment, the stability and reliability of the system are difficult to guarantee.

[0008] Therefore, there is an urgent need for a multimodal electromagnetic navigation-assisted fresh cadaver head transnasal endoscopic anatomical measurement device and method to solve at least one of the above problems. Summary of the Invention

[0009] The present application provides a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device and method for fresh cadaver heads, aiming to solve the problem that traditional navigation systems mainly rely on preoperative imaging data (such as CT and MRI) for positioning, but these systems have limitations such as low image registration accuracy, insufficient real-time feedback, complex operation and safety issues.

[0010] In a first aspect, the present application provides a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for a fresh cadaver head, which is applied to a control module of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head. The multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head further comprises an endoscope module and an electromagnetic tracking module. The front end of the endoscope module is used to extend into the nasal cavity of the fresh cadaver head to be measured. The electromagnetic tracking module is disposed at the front end of the endoscope module and is used to measure magnetic field information of the nasal cavity. The method comprises:

[0011] Acquiring multimodal image information corresponding to a fresh cadaver head to be measured, wherein the multimodal image information includes at least CT image information and MRI image information;

[0012] generating a three-dimensional model corresponding to the fresh cadaver head according to the multimodal image information;

[0013] Obtaining a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of a nasal cavity, a paranasal sinus, a blood vessel, and a nerve;

[0014] When the endoscope module enters the nasal cavity of the fresh cadaver, obtaining magnetic field information sent by the electromagnetic tracking module;

[0015] Calculating the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model according to the magnetic field information;

[0016] generating navigation information corresponding to the endoscope module according to the spatial position and the target position, the navigation information including direction information and distance information;

[0017] The endoscope module is moved to the target position according to the navigation information.

[0018] In some embodiments, generating a three-dimensional model corresponding to the fresh cadaver head based on the multimodal image information includes: performing initial alignment and fine alignment on the CT image information and the MRI image information; inputting the CT image information and the MRI image information into a pre-trained multi-task segmentation model, respectively, completing automatic segmentation of the CT image information and the MRI image information, and obtaining segmentation information corresponding to the CT image information and the MRI image information, respectively; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively; generating multiple label information corresponding to the CT three-dimensional model and the MRI three-dimensional model based on natural language processing technology; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model based on the label information, and obtaining a three-dimensional model corresponding to the fresh cadaver head.

[0019] Exemplarily, before respectively performing initial registration and fine registration on the CT image information and the MRI image information, the method further includes: preprocessing the CT image information and the MRI image information; the preprocessing includes noise removal, standardization and slice alignment.

[0020] Exemplarily, the initial registration and fine registration of the CT image information and the MRI image information respectively include: respectively obtaining feature point information corresponding to the CT image information and the MRI image information; performing a rigid transformation on the CT image information and the MRI image information according to the feature point information to complete the initial registration of the CT image information and the MRI image information; and performing a thin plate spline transformation on the CT image information and the MRI image information to complete the fine registration.

[0021] Exemplarily, the segmentation information includes voxel data, a binary image, and local feature data; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively, includes: obtaining a surface mesh from the binary image; adjusting the density of the surface mesh according to the local feature data; and generating the CT three-dimensional model and the MRI three-dimensional model based on the voxel data and the surface mesh corresponding to the CT image information and the MRI image information, respectively, based on a ray casting algorithm.

[0022] Exemplarily, the label information corresponds to a local area in the CT three-dimensional model or the MRI three-dimensional model; the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the label information, including: obtaining the structure identifier corresponding to each of the label information; determining the fusion weight corresponding to the local area according to the structure information; if the structure identifier is bone, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is greater than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; if the structure identifier is soft tissue, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is less than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; and the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the fusion weight.

[0023] In some embodiments, the electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor, the electromagnetic tracking sensor is arranged at the front end of the endoscope module, and the magnetic field generator is arranged on the outside of the fresh corpse head to generate an external magnetic field; the calculation of the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information includes: obtaining distribution information corresponding to the external magnetic field; and calculating the spatial position based on the distribution information and magnetic field information.

[0024] Exemplarily, calculating the spatial position based on the distribution information and magnetic field information includes: obtaining the generator position and magnetic moment of the magnetic field generator; constructing a magnetic field strength model corresponding to the electromagnetic tracking sensor based on the generator position and magnetic moment of the magnetic field generator; and calculating the spatial position by inversely solving the magnetic field strength model based on the magnetic field information.

[0025] It should be noted that, in some embodiments, the calculating the spatial position by inversely solving the magnetic field strength model according to the magnetic field information includes: obtaining a corresponding magnetic field strength value according to the magnetic field information; and calculating the spatial position by inversely solving the magnetic field strength model according to the distribution information and the magnetic field strength value; the expression of the magnetic field strength model includes:

[0026] ;

[0027] ;

[0028] in, is the magnetic field strength model, is the magnetic field strength value, is the spatial position corresponding to the electromagnetic tracking sensor, is the generator position, is the difference coefficient between the spatial position and the generator position, is the vacuum permeability, is the magnetic moment.

[0029] In a second aspect, the present application provides a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head, comprising:

[0030] an endoscope module, the front end of which is used to extend into the nasal cavity of a fresh cadaver to be measured;

[0031] an electromagnetic tracking module, which is disposed at the front end of the endoscope module and is used to measure magnetic field information of the nasal cavity;

[0032] A control module, the control module comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and, when executing the computer program, implement the steps of the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for a fresh cadaver head as described in the first aspect above.

[0033] This application discloses a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device and method for fresh cadaver heads. The multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for fresh cadaver heads mainly includes the following steps and technical points:

[0034] First, multimodal imaging information is collected from the fresh cadaver head to be measured. This information includes at least CT and MRI images. These two imaging techniques provide complementary information: CT images clearly display skeletal structure, while MRI images have better resolution of soft tissue, blood vessels, and nerves.

[0035] The multimodal image data is then processed by computer to generate a detailed 3D anatomical model. This model serves as the basis for subsequent navigation and intuitively displays the specific location and morphological characteristics of target areas (such as the nasal cavity, sinuses, blood vessels, and nerves).

[0036] Furthermore, key anatomical areas requiring attention, such as specific sinuses, important blood vessels, or sensitive nerves, are identified and marked within the constructed 3D model, and their exact coordinates are recorded. When the endoscope module is inserted into the cadaver's nasal cavity, the electromagnetic tracking device at its front end begins operating, continuously monitoring changes in the surrounding magnetic field and transmitting the collected data to the control system.

[0037] Furthermore, based on the signals received from the electromagnetic sensor, the system can accurately calculate the spatial coordinates of the current front end of the endoscope and map them to the previously established three-dimensional digital model.

[0038] Finally, based on the relative relationship between the current position and the predetermined target, the system automatically generates navigation suggestions, telling the operator which direction to move next and how much distance is needed to reach the destination. Following the provided guidance, the endoscope's position is adjusted until it successfully reaches the designated location and completes the measurement task.

[0039] The proposed method combines multiple imaging techniques to create highly accurate 3D reconstruction models and employs advanced electromagnetic positioning technology for precise spatial localization, significantly enhancing the ability to discern subtle structures during surgery. Compared to traditional methods that rely on static image data, this solution provides immediate feedback during the procedure, helping to avoid the risk of accidental injury to vital organs or tissues. The entire process is highly automated, reducing errors caused by human error and making even complex and delicate procedures easier to master.

[0040] At the same time, for users, the equipment used in the provided methods can not only help them better understand human anatomy knowledge, but also allow them to fully practice before actual operation.

[0041] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0043] Figure 1 This is a schematic flow chart of the steps of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for a fresh cadaver head provided in one embodiment of the present application;

[0044] Figure 2 This is a schematic diagram of a scenario corresponding to the method provided in one embodiment of the present application;

[0045] Figure 3 This is a schematic structural diagram of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head provided in one embodiment of the present application;

[0046] Figure 4 This is a schematic block diagram of the structure of a control module provided in one embodiment of the present application.

[0047] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0049] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0050] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, terms such as "first" and "second" are used to distinguish between identical or similar items having substantially the same functions and effects. Those skilled in the art will understand that terms such as "first" and "second" do not limit the quantity or order of execution, and that terms such as "first" and "second" do not necessarily define differences.

[0051] It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0052] It will also be understood that the term "and / or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0053] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features therein may be combined with each other.

[0054] In modern medicine, particularly in otolaryngology and neurosurgery, endoscopic endoscopic approach (EEA) has become an important minimally invasive surgical technique. EEA, which accesses the skull base through the nasal cavity, can be used to treat a variety of conditions, such as pituitary tumors and meningiomas. However, due to the complex anatomy of the nasal cavity and paranasal sinuses and their significant individual variability, highly precise navigation and positioning systems are required to ensure safety and effectiveness during surgery.

[0055] Traditional navigation systems rely primarily on preoperative imaging data (such as CT and MRI) for positioning, but these systems have the following limitations:

[0056] 1. Low image registration accuracy: The registration error between image data of different modalities is large, affecting navigation accuracy.

[0057] 2. Insufficient real-time feedback: Traditional systems lack real-time dynamic update capabilities and cannot adjust according to changes during the surgical process.

[0058] 3. Complex operation: Key anatomical structures need to be manually marked, which increases the workload of doctors.

[0059] 4. Security issues: In a complex environment, the stability and reliability of the system are difficult to guarantee.

[0060] Therefore, there is an urgent need for a multimodal electromagnetic navigation-assisted fresh cadaver head transnasal endoscopic anatomical measurement device and method to solve at least one of the above problems.

[0061] To resolve the above issues, please refer to Figure 1 , Figure 1This is a schematic flow chart of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for a fresh cadaver head, provided in one embodiment of the present application. This method is applied to the control module of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head. The device also includes an endoscope module and an electromagnetic tracking module. The endoscope module's front end is configured to extend into the nasal cavity of the fresh cadaver head to be measured. The electromagnetic tracking module is located at the front end of the endoscope module and is configured to measure magnetic field information in the nasal cavity.

[0062] To solve the above problems, please refer to Figure 1 Specifically, Figure 1 As shown, the provided method includes steps S101 to S107. The details are as follows:

[0063] S101. Obtain multimodal image information corresponding to a fresh cadaver head to be measured, where the multimodal image information includes at least CT image information and MRI image information.

[0064] Specifically, detailed imaging data of fresh cadaver heads is first obtained using imaging technologies such as CT (computed tomography) and MRI (magnetic resonance imaging). This imaging data provides anatomical information at different levels and types, such as skeletal structure, soft tissue distribution, and the location of blood vessels and nerves. For example, a high-resolution CT scanner is used to obtain detailed bony images of the nasal cavity and surrounding areas; simultaneously, an MRI device is used to obtain soft tissue images of the same area, including sinuses, blood vessels, and nerves. This multimodal imaging information provides more comprehensive and detailed anatomical information, helping to improve the accuracy of 3D model construction in subsequent steps.

[0065] S102. Generate a three-dimensional model corresponding to the fresh corpse head based on the multimodal image information.

[0066] Specifically, the S101 combines imaging data from different modalities, generating a comprehensive 3D anatomical model using medical image processing software or algorithms (e.g., deep learning-based methods). Advanced image registration techniques, for example, align and merge CT and MRI data to create a 3D model encompassing structures such as bones, soft tissue, blood vessels, and nerves. This 3D model provides an intuitive and precise reference for subsequent surgical planning, enhancing surgical safety and accuracy.

[0067] For example, based on CT / MRI DICOM data, a fused 3D model including bones, blood vessels, and nerves can be generated using the Marching Cubes algorithm or a deep learning model (such as 3D U-Net). Mesh smoothing and topology repair are performed using MeshLab. The reconstructed 3D model of the nasopharynx can be interactively rotated and scaled, and the distance between the internal carotid artery and the optic canal can be annotated. This replaces manual labeling, provides a panoramic anatomical view, and reduces the cognitive load on physicians.

[0068] For example, CT bone segmentation uses the threshold method to extract voxels ≥200HU (corresponding to bone tissue). MRI blood vessel segmentation uses the Frangi filter to enhance tubular structures and combines the region growing method to extract blood vessels. Surface reconstruction uses the Marching Cubes algorithm on CT data, and the isosurface threshold is set to 150HU. 3D Residual U-Net is used to perform topological repair on MRI blood vessels, outputting an STL format mesh with a triangular patch size of ≤0.1mm and curvature smoothing iterations of 5 times. The diameter error of the reconstructed three-dimensional model of the internal carotid artery is ≤0.15mm (compared with microanatomical measurements). Multi-tissue fusion modeling is achieved, and the accuracy of blood vessel modeling reaches sub-millimeter level, supporting precise avoidance during surgery.

[0069] S103. Obtain a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of the nasal cavity, paranasal sinuses, blood vessels, and nerves.

[0070] Specifically, on the constructed 3D model, professionals or automated recognition algorithms mark the key anatomical structures of particular interest and their specific coordinates. For example, a doctor manually or with the aid of intelligent tools can mark key points such as the nasal entrance, sinus openings, important vascular pathways, and nerve paths within the 3D model. This clear definition of target locations facilitates more precise surgical planning and ensures that damage to vital organs is avoided during the actual procedure.

[0071] S104. When the endoscope module enters the nasal cavity of the fresh cadaver, the magnetic field information sent by the electromagnetic tracking module is obtained.

[0072] Specifically, once the endoscope's tip is inserted into the patient's nasal cavity, the electromagnetic sensor mounted on it begins operating, continuously monitoring changes in the magnetic field in the surrounding environment and transmitting the collected data to the control system in real time. As the endoscope penetrates deeper into the nasal cavity, the micro-electromagnetic sensor on its tip continuously detects changes in the magnetic field strength and direction in the surrounding space. By monitoring the magnetic field information in real time, the system can accurately determine the current location of the endoscope, thereby achieving dynamic positioning.

[0073] At the same time, if Figure 2 As shown, the method utilizes a flexible navigation probe ( Figure 2(The red part of the probe corresponding to mark 4 in the figure) If you mark the left and right endpoints of the anterior communicating artery, the distance between the probe and the anterior communicating artery will be automatically generated.

[0074] S105. Calculate the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information.

[0075] Specifically, based on the received magnetic field data and a pre-established magnetic field distribution map, a specific algorithm is used to infer the exact coordinates of the electromagnetic tracking device relative to the 3D anatomical model. Assuming the standard magnetic field value B0 at a fixed point A is known, and the actual value currently measured is B1, the difference between the two can be used to infer the relative displacements Δx, Δy, and Δz between the current position P and A. This process maintains a high level of position tracking accuracy even in complex environments, enhancing the overall reliability of the system.

[0076] S106. Generate navigation information corresponding to the endoscope module according to the spatial position and the target position, where the navigation information includes direction information and distance information.

[0077] Specifically, the system compares the endoscope's current position with the previously set target location and generates navigation guidance, including heading and distance information. If the target point is 5 cm to the right, the system outputs the command "turn right 10 degrees and continue forward 5 cm." Clear and concise navigation guidance helps operators reach their destination quickly and accurately, reducing the risks of blind groping.

[0078] Navigation information can be generated by improving the A* algorithm, the cost function is:

[0079] ;

[0080] is the cost function, is the risk weight, risk function Calculated based on blood vessel / nerve proximity, Represents the distance from the starting node to the current node The actual cost (or cost). This value is usually calculated by accumulating the actual cost of each step from the starting point to the current node. Indicates that from the current node Estimated cost to the goal node (heuristic function).

[0081] Indicates the current node The risk value can be used to assess the safety of passing through the node. The risk value can be calculated based on factors such as the proximity of anatomical structures and tissue type. Its expression can include:

[0082] .

[0083] is a node With the distances to key anatomical structures, is a small constant (such as 0.01) that prevents the denominator from being zero.

[0084] S107. Move the endoscope module to the target position according to the navigation information.

[0085] Specifically, the endoscope's posture and trajectory are adjusted according to the navigation suggestions provided in S106 until it reaches the designated anatomical site. Based on the arrow direction and distance values ​​displayed on the screen, the endoscope's angle and length are gradually fine-tuned until its tip contacts the pre-calibrated critical structure. This entire process effectively transitions from static planning to dynamic execution, significantly improving surgical efficiency and success rates.

[0086] In some embodiments, generating a three-dimensional model corresponding to the fresh cadaver head based on the multimodal image information includes: performing initial alignment and fine alignment on the CT image information and the MRI image information; inputting the CT image information and the MRI image information into a pre-trained multi-task segmentation model, respectively, completing automatic segmentation of the CT image information and the MRI image information, and obtaining segmentation information corresponding to the CT image information and the MRI image information, respectively; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively; generating multiple label information corresponding to the CT three-dimensional model and the MRI three-dimensional model based on natural language processing technology; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model based on the label information, and obtaining a three-dimensional model corresponding to the fresh cadaver head.

[0087] Before performing initial and fine registration, the CT and MRI image information is preprocessed. Preprocessing steps include: Noise removal: Using filtering techniques (such as median filtering and Gaussian filtering) to remove noise from the image and improve image quality. Standardization: Adjusting image data from different modalities to a uniform grayscale or intensity range for subsequent processing. Slice alignment: Ensuring that CT and MRI image slices are spatially aligned to avoid registration errors caused by inconsistent slice positions.

[0088] By removing noise, we can reduce interference factors and make the image clearer. Standardization makes data from different modalities comparable, facilitating subsequent registration and fusion. Slice alignment ensures spatial correspondence between images from different modalities, improving registration accuracy.

[0089] Exemplarily, before respectively performing initial registration and fine registration on the CT image information and the MRI image information, the method further includes: preprocessing the CT image information and the MRI image information; the preprocessing includes noise removal, standardization and slice alignment.

[0090] Key feature points, typically salient points in anatomical structures, are extracted from the CT and MRI images. Based on these extracted feature points, rigid transformations (translation and rotation) are performed on the CT and MRI images to complete initial registration. Based on this initial registration, thin plate spline transformation (a nonlinear transformation method) is further used to fine-tune the registration of the CT and MRI images to eliminate errors caused by local deformation.

[0091] By matching feature points and using rigid transformations, we can quickly find the approximate correspondence between the two modal images. Thin plate spline transformations can fine-tune the correspondence of local regions, further improving registration accuracy and generating a more accurate 3D model.

[0092] Exemplarily, the initial registration and fine registration of the CT image information and the MRI image information respectively include: respectively obtaining feature point information corresponding to the CT image information and the MRI image information; performing a rigid transformation on the CT image information and the MRI image information according to the feature point information to complete the initial registration of the CT image information and the MRI image information; and performing a thin plate spline transformation on the CT image information and the MRI image information to complete the fine registration.

[0093] Segmentation information includes voxel data, binary images, and local feature data. A surface mesh is extracted from the binary image to represent the surface of the anatomical structure. The density of the surface mesh is adjusted based on the local feature data to better reflect the details of the anatomical structure. Using a ray casting algorithm, the voxel data and surface mesh are combined to generate a 3D CT model and a 3D MRI model, respectively.

[0094] By combining voxel data with surface meshes, more detailed and accurate 3D models can be generated. Adjusting the mesh density allows for optimization based on the importance of different areas, enhancing the detail of the model. Ray casting is an efficient 3D reconstruction method that can generate high-quality 3D models in a short time.

[0095] Exemplarily, the segmentation information includes voxel data, a binary image, and local feature data; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively, includes: obtaining a surface mesh from the binary image; adjusting the density of the surface mesh according to the local feature data; and generating the CT three-dimensional model and the MRI three-dimensional model based on the voxel data and the surface mesh corresponding to the CT image information and the MRI image information, respectively, based on a ray casting algorithm.

[0096] Segmentation information includes voxel data, binary images, and local feature data. A surface mesh is extracted from the binary image to represent the surface of the anatomical structure. The density of the surface mesh is adjusted based on the local feature data to better reflect the details of the anatomical structure. Using a ray casting algorithm, the voxel data and surface mesh are combined to generate a 3D CT model and a 3D MRI model, respectively.

[0097] By combining voxel data with surface meshes, more detailed and accurate 3D models can be generated. Adjusting the mesh density allows for optimization based on the importance of different areas, enhancing the detail of the model. Ray casting is an efficient 3D reconstruction method that can generate high-quality 3D models in a short time.

[0098] Exemplarily, the label information corresponds to a local area in the CT three-dimensional model or the MRI three-dimensional model; the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the label information, including: obtaining the structure identifier corresponding to each of the label information; determining the fusion weight corresponding to the local area according to the structure information; if the structure identifier is bone, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is greater than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; if the structure identifier is soft tissue, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is less than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; and the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the fusion weight.

[0099] The label information corresponds to a local region in the CT 3D model or MRI 3D model. The structural identifier (e.g., bone, soft tissue, etc.) corresponding to each label information is obtained. If the structural identifier is bone, the fusion weight corresponding to the local region in the CT 3D model is greater than the fusion weight corresponding to the local region in the MRI 3D model. If the structural identifier is soft tissue, the fusion weight corresponding to the local region in the CT 3D model is less than the fusion weight corresponding to the local region in the MRI 3D model. Based on the fusion weights, the CT 3D model and the MRI 3D model are fused to generate a comprehensive 3D model.

[0100] By using different fusion weights, the strengths of each modality can be better preserved, such as CT's advantages in bone imaging and MRI's advantages in soft tissue imaging. The comprehensive 3D model provides more comprehensive and detailed anatomical information, helping to improve the accuracy of surgical planning and navigation. The fused model can better meet clinical needs and provide a more reliable and practical reference.

[0101] In some embodiments, the electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor, the electromagnetic tracking sensor is arranged at the front end of the endoscope module, and the magnetic field generator is arranged on the outside of the fresh corpse head to generate an external magnetic field; the calculation of the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information includes: obtaining distribution information corresponding to the external magnetic field; and calculating the spatial position based on the distribution information and magnetic field information.

[0102] The electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor. The specific configuration is as follows: The magnetic field generator is located outside the fresh cadaver head and generates an external magnetic field. The electromagnetic tracking sensor is located at the front end of the endoscope module and measures the magnetic field information within the nasal cavity.

[0103] The known position and magnetic moment of the magnetic field generator are used to construct a distribution model of the external magnetic field. Using this magnetic field distribution information and the actual measured magnetic field information, the precise position of the electromagnetic tracking sensor in the 3D model is calculated.

[0104] The combination of an external magnetic field generator and internal sensors enables high-precision spatial positioning. This system monitors the endoscope's position in real time, providing dynamic updates and improving surgical navigation accuracy. This reduces the workload of manual marking and adjustment, improving surgical efficiency and safety.

[0105] Exemplarily, calculating the spatial position based on the distribution information and magnetic field information includes: obtaining the generator position and magnetic moment of the magnetic field generator; constructing a magnetic field strength model corresponding to the electromagnetic tracking sensor based on the generator position and magnetic moment of the magnetic field generator; and calculating the spatial position by inversely solving the magnetic field strength model based on the magnetic field information.

[0106] The specific location and magnetic moment of the magnetic field generator are determined. Based on this location and magnetic moment, a magnetic field strength model corresponding to the electromagnetic tracking sensor is constructed. Based on the actual measured magnetic field information, the spatial position of the electromagnetic tracking sensor is calculated by inversely solving the magnetic field strength model.

[0107] By constructing a detailed magnetic field strength model, we can more accurately describe the magnetic field distribution and improve positioning accuracy. The inverse solution method can infer the specific position of the sensor from the actual measurement value, ensuring positioning accuracy. This method supports real-time calculation, enabling dynamic adjustment of the endoscope's position during surgery, improving navigation flexibility and reliability.

[0108] It should be noted that, in some embodiments, the calculating the spatial position by inversely solving the magnetic field strength model according to the magnetic field information includes: obtaining a corresponding magnetic field strength value according to the magnetic field information; and calculating the spatial position by inversely solving the magnetic field strength model according to the distribution information and the magnetic field strength value; the expression of the magnetic field strength model includes:

[0109] ;

[0110] ;

[0111] in, is the magnetic field strength model, is the magnetic field strength value, is the spatial position corresponding to the electromagnetic tracking sensor, is the generator position, is the difference coefficient between the spatial position and the generator position, is the vacuum permeability, is the magnetic moment.

[0112] The electromagnetic tracking sensor obtains the actual measured magnetic field strength value. Based on the magnetic field distribution information and the actual measured magnetic field strength value, the spatial position of the electromagnetic tracking sensor is calculated by inverse solving the magnetic field strength model. Using a specific mathematical model, the spatial position of the electromagnetic tracking sensor can be accurately calculated. This inverse solution method based on the magnetic field strength model enables high-precision spatial positioning and reduces errors. This method supports real-time calculations, enabling dynamic adjustment of the endoscope's position during surgery, improving the real-time and accuracy of navigation. The use of a mathematical model makes the positioning process more stable and reliable, reducing the impact of environmental factors on the positioning results.

[0113] See also Figure 3 As shown, Figure 31 is a schematic diagram of the structure of a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device 200 for a fresh cadaver head, provided in an embodiment of the present application. The multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device 200 is used to perform the steps of the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for a fresh cadaver head shown in the above-described embodiments. The multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device 200 can be a single server or a server cluster, or the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device 200 can be a terminal, such as a handheld terminal, a laptop computer, a wearable device, or a robot.

[0114] like Figure 3 As shown, the multimodal electromagnetic navigation-assisted fresh cadaver head transnasal endoscopic anatomical measurement device 200 includes:

[0115] The multimodal acquisition unit 201 is configured to acquire multimodal image information corresponding to a fresh cadaver head to be measured, wherein the multimodal image information includes at least CT image information and MRI image information;

[0116] A three-dimensional generation unit 202 is configured to generate a three-dimensional model corresponding to the fresh cadaver head based on the multimodal image information;

[0117] A position acquisition unit 203 is configured to acquire a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of the nasal cavity, paranasal sinuses, blood vessels, and nerves;

[0118] an information acquisition unit 204, configured to acquire magnetic field information sent by the electromagnetic tracking module when the endoscope module enters the nasal cavity of the fresh cadaver;

[0119] A spatial acquisition unit 205 is configured to calculate a spatial position corresponding to the electromagnetic tracking module in the three-dimensional model according to the magnetic field information;

[0120] a navigation generating unit 206, configured to generate navigation information corresponding to the endoscope module according to the spatial position and the target position, the navigation information including direction information and distance information;

[0121] The target moving unit 207 is configured to move the endoscope module to the target position according to the navigation information.

[0122] In some embodiments, generating a three-dimensional model corresponding to the fresh cadaver head based on the multimodal image information includes: performing initial alignment and fine alignment on the CT image information and the MRI image information; inputting the CT image information and the MRI image information into a pre-trained multi-task segmentation model, respectively, completing automatic segmentation of the CT image information and the MRI image information, and obtaining segmentation information corresponding to the CT image information and the MRI image information, respectively; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively; generating multiple label information corresponding to the CT three-dimensional model and the MRI three-dimensional model based on natural language processing technology; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model based on the label information, and obtaining a three-dimensional model corresponding to the fresh cadaver head.

[0123] Exemplarily, before respectively performing initial registration and fine registration on the CT image information and the MRI image information, the method further includes: preprocessing the CT image information and the MRI image information; the preprocessing includes noise removal, standardization and slice alignment.

[0124] Exemplarily, the initial registration and fine registration of the CT image information and the MRI image information respectively include: respectively obtaining feature point information corresponding to the CT image information and the MRI image information; performing a rigid transformation on the CT image information and the MRI image information according to the feature point information to complete the initial registration of the CT image information and the MRI image information; and performing a thin plate spline transformation on the CT image information and the MRI image information to complete the fine registration.

[0125] Exemplarily, the segmentation information includes voxel data, a binary image, and local feature data; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively, includes: obtaining a surface mesh from the binary image; adjusting the density of the surface mesh according to the local feature data; and generating the CT three-dimensional model and the MRI three-dimensional model based on the voxel data and the surface mesh corresponding to the CT image information and the MRI image information, respectively, based on a ray casting algorithm.

[0126] Exemplarily, the label information corresponds to a local area in the CT three-dimensional model or the MRI three-dimensional model; the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the label information, including: obtaining the structure identifier corresponding to each of the label information; determining the fusion weight corresponding to the local area according to the structure information; if the structure identifier is bone, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is greater than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; if the structure identifier is soft tissue, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is less than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; and the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the fusion weight.

[0127] In some embodiments, the electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor, the electromagnetic tracking sensor is arranged at the front end of the endoscope module, and the magnetic field generator is arranged on the outside of the fresh corpse head to generate an external magnetic field; the calculation of the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information includes: obtaining distribution information corresponding to the external magnetic field; and calculating the spatial position based on the distribution information and magnetic field information.

[0128] Exemplarily, calculating the spatial position based on the distribution information and magnetic field information includes: obtaining the generator position and magnetic moment of the magnetic field generator; constructing a magnetic field strength model corresponding to the electromagnetic tracking sensor based on the generator position and magnetic moment of the magnetic field generator; and calculating the spatial position by inversely solving the magnetic field strength model based on the magnetic field information.

[0129] It should be noted that, in some embodiments, the calculating the spatial position by inversely solving the magnetic field strength model according to the magnetic field information includes: obtaining a corresponding magnetic field strength value according to the magnetic field information; and calculating the spatial position by inversely solving the magnetic field strength model according to the distribution information and the magnetic field strength value; the expression of the magnetic field strength model includes:

[0130] ;

[0131] ;

[0132] in, is the magnetic field strength model, is the magnetic field strength value, is the spatial position corresponding to the electromagnetic tracking sensor, is the generator position, is the difference coefficient between the spatial position and the generator position, is the vacuum permeability, is the magnetic moment.

[0133] It should be noted that, those skilled in the art can clearly understand that, for the sake of convenience and brevity of description, the specific working processes of the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head and each unit described above can refer to the corresponding processes in the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement embodiments of a fresh cadaver head described in the above embodiments, and will not be repeated here.

[0134] The above-mentioned multimodal electromagnetic navigation-assisted endoscopic anatomical measurement of fresh cadaver head can be realized in the form of a computer program. The computer program can be used in the following ways: Figure 3 Run on the device shown.

[0135] An embodiment of the present application provides a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head, comprising: an endoscope module, the front end of which is used to extend into the nasal cavity of a fresh cadaver head to be measured; an electromagnetic tracking module, the electromagnetic tracking module being arranged at the front end of the endoscope module and being used to measure the magnetic field information of the nasal cavity; a control module, the control module comprising a memory and a processor; the memory being used to store a computer program; and the processor being used to execute the computer program and implement any of the methods described above when executing the computer program.

[0136] The endoscope module's front end is designed to extend into the nasal cavity of a fresh cadaver. It's typically equipped with a high-definition camera and lighting system to provide clear visual feedback during surgery. The module is used to observe the internal structure of the nasal cavity and transmits real-time images to the control module via the camera.

[0137] The electromagnetic tracking module, typically a small sensor, is located at the front end of the endoscope module. It measures magnetic field information within the nasal cavity. This magnetic field is generated by an external magnetic field generator, and the sensor determines its position in three-dimensional space by detecting this magnetic field information.

[0138] The control module stores computer programs, multimodal imaging data, 3D models, and other information. The processor executes the computer programs stored in the memory to perform the following functions: It acquires multimodal imaging information (such as CT and MRI) and generates a 3D model. It also marks the target locations of key anatomical structures within the 3D model. It calculates their spatial positions within the 3D model based on the magnetic field information transmitted by the electromagnetic tracking module. It also generates navigation information, including direction and distance information. It also controls the movement of the endoscope module to the target location.

[0139] For example, a high-resolution CT scanner is used to obtain detailed bony images of the nasal cavity and surrounding areas, while an MRI device is used to obtain soft tissue images of the same area. The CT and MRI image data are preprocessed, including noise removal, normalization, and slice alignment, followed by initial and fine registration.

[0140] Feature point information is extracted from CT and MRI images, and a rigid transformation is performed for initial registration. Thin plate spline transformation is then used for fine registration. Based on the segmentation information (voxel data, binary images, and local feature data), CT and MRI 3D models are generated, and finally fused to create a comprehensive 3D model.

[0141] An external magnetic field generator generates a magnetic field with a known distribution, and an electromagnetic tracking sensor at the endoscope's tip measures the actual magnetic field. A magnetic field strength model is constructed based on the generator's position and magnetic moment, and the sensor's spatial position is calculated through inverse engineering. Navigation information, including direction and distance, is generated based on the sensor's spatial position and the pre-marked locations of key anatomical structures. The endoscope's posture and trajectory are adjusted based on this navigation information until the target location is reached.

[0142] The three-dimensional model generated by multimodal imaging data and electromagnetic tracking technology can achieve high-precision spatial positioning and reduce errors during surgery. The electromagnetic tracking module can monitor the position changes of the endoscope in real time, provide dynamic update capabilities, and ensure the timeliness and accuracy of navigation information. Automated image processing and navigation information generation reduce the workload of manual marking and adjustment, improving operational efficiency. High-precision positioning and real-time feedback help avoid damage to vital organs and improve the safety and success rate of surgery. Multimodal imaging data provides more comprehensive and detailed anatomical structure information, which helps to formulate more accurate operation plans. By dynamically adjusting the position of the endoscope, the stability and reliability of the system can be maintained in complex environments to adapt to different surgical needs.

[0143] This multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for fresh cadaver heads achieves high-precision, real-time feedback, and automated surgical navigation through the collaborative operation of an endoscopic module, an electromagnetic tracking module, and a control module. This device not only improves surgical safety and success rates, but also simplifies the procedure and reduces the workload of surgeons, demonstrating its broad potential for application.

[0144] See also Figure 4 , Figure 4 1 is a schematic block diagram of the structure of a control module provided in an embodiment of the present application. The control module includes a processor, a memory and a network interface connected via a device bus, wherein the memory may include a storage medium and an internal memory.

[0145] The storage medium can store an operating device and a computer program. The computer program includes program instructions, which, when executed, can enable the processor to perform any multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement of a fresh cadaver head.

[0146] The processor is used to provide computing and control capabilities and support the operation of the entire control module.

[0147] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can perform any multi-modal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement of a fresh cadaver head.

[0148] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the terminal to which the solution of the present application is applied. The specific control module may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0149] It should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0150] In one embodiment, the processor is configured to execute a computer program stored in the memory to implement the following steps:

[0151] Acquiring multimodal image information corresponding to a fresh cadaver head to be measured, wherein the multimodal image information includes at least CT image information and MRI image information;

[0152] generating a three-dimensional model corresponding to the fresh cadaver head according to the multimodal image information;

[0153] Obtaining a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of a nasal cavity, a paranasal sinus, a blood vessel, and a nerve;

[0154] When the endoscope module enters the nasal cavity of the fresh cadaver, obtaining magnetic field information sent by the electromagnetic tracking module;

[0155] Calculating the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model according to the magnetic field information;

[0156] generating navigation information corresponding to the endoscope module according to the spatial position and the target position, the navigation information including direction information and distance information;

[0157] The endoscope module is moved to the target position according to the navigation information.

[0158] In some embodiments, generating a three-dimensional model corresponding to the fresh cadaver head based on the multimodal image information includes: performing initial alignment and fine alignment on the CT image information and the MRI image information; inputting the CT image information and the MRI image information into a pre-trained multi-task segmentation model, respectively, completing automatic segmentation of the CT image information and the MRI image information, and obtaining segmentation information corresponding to the CT image information and the MRI image information, respectively; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively; generating multiple label information corresponding to the CT three-dimensional model and the MRI three-dimensional model based on natural language processing technology; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model based on the label information, and obtaining a three-dimensional model corresponding to the fresh cadaver head.

[0159] Exemplarily, before respectively performing initial registration and fine registration on the CT image information and the MRI image information, the method further includes: preprocessing the CT image information and the MRI image information; the preprocessing includes noise removal, standardization and slice alignment.

[0160] Exemplarily, the initial registration and fine registration of the CT image information and the MRI image information respectively include: respectively obtaining feature point information corresponding to the CT image information and the MRI image information; performing a rigid transformation on the CT image information and the MRI image information according to the feature point information to complete the initial registration of the CT image information and the MRI image information; and performing a thin plate spline transformation on the CT image information and the MRI image information to complete the fine registration.

[0161] Exemplarily, the segmentation information includes voxel data, a binary image, and local feature data; generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively, includes: obtaining a surface mesh from the binary image; adjusting the density of the surface mesh according to the local feature data; and generating the CT three-dimensional model and the MRI three-dimensional model based on the voxel data and the surface mesh corresponding to the CT image information and the MRI image information, respectively, based on a ray casting algorithm.

[0162] Exemplarily, the label information corresponds to a local area in the CT three-dimensional model or the MRI three-dimensional model; the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the label information, including: obtaining the structure identifier corresponding to each of the label information; determining the fusion weight corresponding to the local area according to the structure information; if the structure identifier is bone, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is greater than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; if the structure identifier is soft tissue, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is less than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; and the fusion of the CT three-dimensional model and the MRI three-dimensional model is completed according to the fusion weight.

[0163] In some embodiments, the electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor, the electromagnetic tracking sensor is arranged at the front end of the endoscope module, and the magnetic field generator is arranged on the outside of the fresh corpse head to generate an external magnetic field; the calculation of the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model based on the magnetic field information includes: obtaining distribution information corresponding to the external magnetic field; and calculating the spatial position based on the distribution information and magnetic field information.

[0164] Exemplarily, calculating the spatial position based on the distribution information and magnetic field information includes: obtaining the generator position and magnetic moment of the magnetic field generator; constructing a magnetic field strength model corresponding to the electromagnetic tracking sensor based on the generator position and magnetic moment of the magnetic field generator; and calculating the spatial position by inversely solving the magnetic field strength model based on the magnetic field information.

[0165] It should be noted that, in some embodiments, the calculating the spatial position by inversely solving the magnetic field strength model according to the magnetic field information includes: obtaining a corresponding magnetic field strength value according to the magnetic field information; and calculating the spatial position by inversely solving the magnetic field strength model according to the distribution information and the magnetic field strength value; the expression of the magnetic field strength model includes:

[0166] ;

[0167] ;

[0168] in, is the magnetic field strength model, is the magnetic field strength value, is the spatial position corresponding to the electromagnetic tracking sensor, is the generator position, is the difference coefficient between the spatial position and the generator position, is the vacuum permeability, is the magnetic moment.

[0169] It should be noted that those skilled in the art can clearly understand that, for the sake of convenience and brevity of description, the specific working process of the processor described above can refer to the corresponding process in the multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement of fresh cadaver heads described in the above embodiments, and will not be repeated here.

[0170] A computer-readable storage medium is also provided in an embodiment of the present application, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and the processor executes the program instructions to implement the steps of the multimodal electromagnetic navigation-assisted fresh cadaver head transnasal endoscopic anatomical measurement method provided in the above-mentioned embodiments of the present application.

[0171] The computer-readable storage medium may be an internal storage unit of the control module described in the aforementioned embodiment, such as a hard disk or memory of the control module. The computer-readable storage medium may also be an external storage device of the control module, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the control module.

[0172] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement method for fresh cadaver heads, characterized in that: The method is applied to a multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head. The device includes a control module, an endoscope module, and an electromagnetic tracking module. The front end of the endoscope module is used to extend into the nasal cavity of the fresh cadaver head to be measured. The electromagnetic tracking module is arranged at the front end of the endoscope module and is used to measure the magnetic field information of the nasal cavity. The method comprises: Acquiring multimodal image information corresponding to a fresh cadaver head to be measured, wherein the multimodal image information includes at least CT image information and MRI image information; The three-dimensional model corresponding to the fresh corpse head is generated according to the multimodal image information, including: performing initial registration and fine registration on the CT image information and the MRI image information; inputting the CT image information and the MRI image information into a pre-trained multi-task segmentation model respectively to complete automatic segmentation of the CT image information and the MRI image information, and obtaining segmentation information corresponding to the CT image information and the MRI image information respectively; generating a CT three-dimensional model and an MRI three-dimensional model according to the segmentation information corresponding to the CT image information and the MRI image information respectively; generating a plurality of label information corresponding to the CT three-dimensional model and the MRI three-dimensional model respectively based on natural language processing technology; the label information is stored in the CT three-dimensional model or the MRI three-dimensional model. I three-dimensional model corresponds to a local area; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model according to the label information, including: obtaining a structure identifier corresponding to each of the label information; determining a fusion weight corresponding to the local area according to the structure identifier; if the structure identifier is bone, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is greater than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; if the structure identifier is soft tissue, the fusion weight corresponding to the local area of ​​the CT three-dimensional model is less than the fusion weight corresponding to the local area of ​​the MRI three-dimensional model; completing the fusion of the CT three-dimensional model and the MRI three-dimensional model according to the fusion weight; obtaining a three-dimensional model corresponding to the fresh cadaver head; Obtaining a target position corresponding to a preset key anatomical structure in the three-dimensional model; the key anatomical structure includes at least one or more of a nasal cavity, a paranasal sinus, a blood vessel, and a nerve; When the endoscope module enters the nasal cavity of the fresh cadaver, obtaining magnetic field information sent by the electromagnetic tracking module; Calculating the corresponding spatial position of the electromagnetic tracking module in the three-dimensional model according to the magnetic field information; generating navigation information corresponding to the endoscope module according to the spatial position and the target position, the navigation information including direction information and distance information; The endoscope module is moved to the target position according to the navigation information.

2. The method according to claim 1, characterized in that Before respectively performing initial registration and fine registration on the CT image information and the MRI image information, the method further includes: The CT image information and the MRI image information are preprocessed; the preprocessing includes noise removal, standardization and slice alignment.

3. The method according to claim 1, characterized in that The performing initial registration and fine registration on the CT image information and the MRI image information respectively includes: Respectively obtaining feature point information corresponding to the CT image information and the MRI image information; Performing a rigid transformation on the CT image information and the MRI image information according to the feature point information to complete initial registration of the CT image information and the MRI image information; The CT image information and the MRI image information are subjected to thin plate spline transformation to complete the fine registration.

4. The method according to claim 1, wherein The segmentation information includes voxel data, a binary image, and local feature data; and generating a CT three-dimensional model and an MRI three-dimensional model based on the segmentation information corresponding to the CT image information and the MRI image information, respectively, includes: obtaining a surface mesh from the binary image; adjusting the density of the surface mesh according to the local feature data; The CT three-dimensional model and the MRI three-dimensional model are generated based on the voxel data and surface mesh corresponding to the CT image information and the MRI image information respectively based on a ray casting algorithm.

5. The method according to claim 1, wherein The electromagnetic tracking module includes a magnetic field generator and an electromagnetic tracking sensor. The electromagnetic tracking sensor is arranged at the front end of the endoscope module, and the magnetic field generator is arranged outside the fresh cadaver head to generate an external magnetic field. Calculating the spatial position corresponding to the electromagnetic tracking module in the three-dimensional model according to the magnetic field information includes: Obtaining distribution information corresponding to the external magnetic field; The spatial position is calculated according to the distribution information and the magnetic field information.

6. The method according to claim 5, characterized in that Calculating the spatial position according to the distribution information and the magnetic field information includes: obtaining a generator position and a magnetic moment of the magnetic field generator; Constructing a magnetic field intensity model corresponding to the electromagnetic tracking sensor according to the generator position and magnetic moment of the magnetic field generator; The spatial position is calculated by reversely solving the magnetic field strength model according to the magnetic field information.

7. The method according to claim 6, characterized in that The calculating the spatial position by inversely solving the magnetic field strength model according to the magnetic field information includes: Acquire a corresponding magnetic field intensity value according to the magnetic field information; The spatial position is calculated by reversely solving the magnetic field strength model according to the distribution information and the magnetic field strength value; the expression of the magnetic field strength model includes: ； ； in, is the magnetic field strength model, is the magnetic field strength value, is the spatial position corresponding to the electromagnetic tracking sensor, is the generator position, is the difference coefficient between the spatial position and the generator position, is the vacuum permeability, is the magnetic moment.

8. A multimodal electromagnetic navigation-assisted transnasal endoscopic anatomical measurement device for a fresh cadaver head for implementing the method according to any one of claims 1 to 7, characterized in that: include: an endoscope module, the front end of which is used to extend into the nasal cavity of a fresh cadaver to be measured; an electromagnetic tracking module, which is disposed at the front end of the endoscope module and is used to measure magnetic field information of the nasal cavity; A control module includes a memory and a processor; the memory is used to store computer programs; and the processor is used to execute the computer programs.

Images