A rapid registration method and system for mixed reality navigation in orthopedic surgery

Abstract

This invention relates to the field of computer-aided surgical technology and discloses a rapid registration method and system for mixed reality navigation in orthopedic surgery. The method includes: calculating an initial alignment transformation matrix between the virtual surgical planning model and the patient's actual anatomical structure based on a physical registration device fixed intraoperatively to the target anatomical site of the patient and a pre-stored spatial relationship between the device and a virtual surgical planning model; calculating a fine registration transformation matrix for optimizing alignment accuracy by collecting spatial point data of the patient's actual anatomical surface and performing algorithmic registration with the surface data of the virtual model; and superimposing and fusing the virtual surgical planning model and its associated preoperative planning parameters into the real field of view of a mixed reality display device to generate and update guidance instructions for guiding the operation of surgical instruments in real time. Through the above-mentioned coarse registration plus fine registration two-stage registration method, high-precision and fast intraoperative real-time fusion navigation is achieved.

Description

Technical Field

[0001] This invention relates to the field of computer-aided surgical technology, and in particular to a rapid registration method and system for mixed reality navigation in orthopedic surgery. Background Technology

[0002] Mixed Reality (MR) surgical navigation provides surgeons with intuitive intraoperative 3D guidance by overlaying virtual surgical planning information onto the surgeon's view of the real surgical field in real time and with high precision. It is hailed as a crucial development direction for next-generation surgical navigation. In complex orthopedic surgeries such as total hip arthroplasty (THA), it is significant for improving the accuracy of acetabular cup implantation and reducing reliance on surgeon experience. However, achieving precise, reliable, and efficient virtual-real fusion is the core challenge in the clinical translation of MR navigation. The key lies in registration, that is, establishing a spatial correspondence between the virtual 3D model and the patient's actual anatomical structure.

[0003] Currently, registration schemes for mixed reality surgical navigation mainly face the following technical bottlenecks: First, traditional registration methods are inefficient in deep surgical areas with limited exposure. For example, manual registration based on anatomical landmarks is a classic method in optical navigation, requiring the surgeon to use a probe to sequentially pinpoint a predefined set of corresponding points on the patient's bones and the virtual model. However, in THA surgery, the acetabulum is a deep, concave hemispherical region with limited surgical field exposure, making it extremely difficult to accurately and quickly identify and pinpoint these landmarks. This process is not only time-consuming but also prone to registration failure or decreased accuracy due to pinpointing errors, soft tissue obstruction, or blood contamination, severely impacting the smoothness of the surgery.

[0004] Secondly, while pure algorithmic registration schemes can avoid manual point selection, they are sensitive to initial pose and computationally time-consuming, making it difficult to meet the real-time requirements of intraoperative procedures. Automatic registration algorithms such as Iterative Closest Point (ICP) can achieve high-precision registration by matching the actual acquired point cloud with the virtual model surface. However, these algorithms have two inherent drawbacks: first, their convergence and final accuracy heavily depend on the quality of the initial pose; if the initial deviation is large, they are prone to getting trapped in local optima; second, the ICP algorithm itself has high computational complexity, making it a computationally intensive task. When running on mobile mixed reality devices such as HoloLens, their limited local computing power often results in a single registration taking several minutes, which is unacceptable in the time-sensitive process of surgery.

[0005] Especially when a surgical guide needs to be customized for each case, registration becomes more difficult, and the planning parameters cannot be adjusted in real time during the operation, resulting in poor flexibility.

[0006] Therefore, current mixed reality surgical navigation technology, especially in scenarios with deep surgical characteristics such as THA, has not yet been able to effectively resolve the contradiction between registration speed, accuracy, and clinical applicability. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a rapid registration method and system for mixed reality navigation in orthopedic surgery, which can quickly provide excellent initial pose and efficiently complete high-precision final registration under limited computing power.

[0008] On the one hand, the present invention provides a rapid registration method for mixed reality navigation in orthopedic surgery, comprising the following steps: S1: An optically recognizable mark is set on a physical registration device that is pre-fixed to the target anatomical site of the patient's bone. The mark is identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. S2: After completing the coarse registration, a probe whose spatial position is tracked in real time is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient, forming an actual surface point cloud; the surface three-dimensional point set of the virtual surgical planning model corresponding to the target bone anatomical site is obtained as the target point cloud; the initial alignment transformation matrix is ​​used as the initial transformation estimate of the iterative nearest point ICP algorithm to perform registration calculation on the actual surface point cloud and the target point set, and the optimized fine registration transformation matrix is ​​obtained to complete the fine registration; S3: Based on the fine registration transformation matrix, the virtual surgical planning model is superimposed and displayed in real time in the field of view of the mixed reality display device, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

[0009] Further, in step S1, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real-world coordinate system is calculated using the following formula: , in, The optical marker coordinate system of the physical registration device, determined during the preoperative planning phase. Relative to the coordinate system of the virtual surgical planning model The initial alignment transformation matrix; Intraoperative time The optical marker coordinate system obtained through tracking by the mixed reality display device The world coordinate system perceived by the mixed reality display device The real-time transformation matrix; Intraoperative time From the patient coordinate system to the optical marker coordinate system The real-time transformation matrix; Intraoperative time Obtained patient coordinate system Relative to the world coordinate system The real-time transformation matrix; The coordinate system of the virtual surgical planning model is calculated. Relative to the patient coordinate system The initial alignment transformation matrix, i.e., the coarse registration result.

[0010] Further, in step S2, based on the coarse registration result, the iterative nearest point (ICP) algorithm is used to perform the fine registration, specifically including: The set of surface vertices of the target region in the preoperative 3D model is used as the target point cloud, where the first vertex is... The coordinates of each point are represented as follows: ; Multiple source point clouds were acquired on the surface of the target anatomical structure corresponding to the patient using a positioning probe, among which the first... The coordinates of each point are calculated using the following formula: , in, To collect the first At that moment, For a moment From the patient coordinate system To the world coordinate system The inverse of the transformation matrix, For a moment From the probe coordinate system To the world coordinate system The transformation matrix, The first probe acquisition The coordinates of the points; The rigid transformation matrix between the source point cloud and the target point cloud is calculated using the Iterative Closest Point (ICP) algorithm. Then the coordinate system of the virtual surgical planning model To the patient coordinate system The final fine registration transformation matrix is: ; To achieve real-time dynamic overlay of the simulated surgical planning model and the actual anatomical structure, the system calculates the coordinate system of the virtual surgical planning model in real time. To the coordinate system of mixed reality display devices Transformation matrix : , in, The coordinates obtained in real time by the augmented reality device through the SLAM self-localization system from the coordinate system of the mixed reality display device. To the world coordinate system The transformation matrix, For a moment From the patient coordinate system To the world coordinate system The transformation matrix.

[0011] Further, in step S3, the generation of the guidance instruction specifically includes: Establish a local anatomical reference system corresponding to the virtual surgical planning model The reference frame It consists of three orthogonal unit basis vectors. , and definition; Using the anatomical forward tilt angle located in the coordinate plane and imaging abduction angle Parameterized definition of the wear direction vector The calculation formula is: , in, The value of is 1 or -1, which is a sign parameter used to distinguish the spatial orientation of the target anatomical structure; Based on the precise registration transformation matrix, the wear direction vector is... Transform to real-world coordinate system Below, the actual guiding direction used for navigation display is obtained. : , in, This represents the overall transformation function.

[0012] Preferably, the guidance instruction further includes intraoperative dynamic adjustment of guidance parameters, including: During navigation, the system receives adjustment commands input by the user, including adjustments to the tilt angle. and abduction angle fine-tuning amount and ; The world coordinate system is updated in real time based on the aforementioned fine-tuning amount. The friction direction vector below The updated formula is: ; The virtual abrasion direction guide displayed on the mixed reality display device is recalculated and updated based on the updated direction vector.

[0013] Furthermore, in step S2, the computation task of the Iterative Closest Point (ICP) algorithm in the fine registration is performed by a remote computing node that is communicatively connected to the mixed reality display device.

[0014] Preferably, the physical registration device is a modular, individualized surgical guide, comprising a fitting component that is individually matched to the patient's skeletal anatomy and a navigation component with the optical markers. The fitting component and the navigation component are detachably connected, and the initial alignment transformation relationship is... The connection between the two and the virtual surgical planning model are determined before the operation.

[0015] On the other hand, the present invention provides a rapid registration system for mixed reality navigation in orthopedic surgery, comprising: The coarse registration module is used to set optically recognizable marks on a physical registration device that is pre-fixed to the target anatomical site of the patient's bones. The marks are then identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. The fine registration module is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient using a probe whose spatial position is tracked in real time after the coarse registration is completed, forming an actual surface point cloud; to obtain the surface three-dimensional point set of the virtual surgical planning model corresponding to the target bone anatomical site as the target point cloud; and to perform registration calculation on the actual surface point cloud and the target point set using the initial alignment transformation matrix as the initial transformation estimate of the iterative nearest point ICP algorithm, so as to obtain the optimized fine registration transformation matrix and complete the fine registration. The fusion navigation module is used to overlay and display the virtual surgical planning model in real time on the field of view of the mixed reality display device according to the fine registration transformation matrix, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

[0016] In addition, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the rapid registration method for mixed reality navigation in orthopedic surgery as described above.

[0017] Meanwhile, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the rapid registration method for mixed reality navigation in orthopedic surgery as described above.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention completes coarse registration within seconds by setting up a physical registration device (such as an individualized guide plate), providing the system with an initial alignment transformation matrix that is close to reality. This lays the iterative starting point for the subsequent fine registration algorithm, enabling it to converge quickly and avoiding lengthy iterations or getting stuck in local optima due to poor initial pose. Based on coarse registration, this invention performs fine registration by matching the actual collected point cloud data. It adaptively fine-tunes and compensates for minor deviations caused by individual differences, intraoperative deformation, or initial installation, thereby ensuring the ultra-high accuracy and clinical reliability of the final registration result. The method of this invention does not rely on anatomical landmarks that are difficult to identify precisely during surgery. Through a strategy of macroscopic physical fixation and microscopic algorithm correction, it effectively overcomes the interference of complex situations such as deep surgical fields, soft tissue obstruction, and intraoperative bleeding on registration, and has stronger robustness. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a rapid registration method for mixed reality navigation in orthopedic surgery according to the present invention; Figure 2 This is a schematic diagram of a registration system structure based on two stages of coarse registration and fine registration according to the present invention; Figure 3 This is a schematic diagram of the coordinate transformation relationship in the registration process of the present invention; Figure 4 This is a schematic diagram of a coarse registration process based on a guide plate according to the present invention; Figure 5 This is a schematic diagram illustrating the process of determining a local pelvic reference system and the direction of abrasion according to the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This embodiment uses the registration and navigation of the acetabular side in total hip arthroplasty (THA) as an example, but the present invention is not limited to this application scenario.

[0022] The specific embodiments of the present invention will be described below with reference to the accompanying drawings and examples.

[0023] Example 1

[0024] Please see Figure 1 The technical solution for a rapid registration method of mixed reality navigation in orthopedic surgery provided in this embodiment includes the following steps: S1: An optically recognizable mark is set on a physical registration device that is pre-fixed to the target anatomical site of the patient's bone. The mark is identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. S2: After completing the coarse registration, a probe whose spatial position is tracked in real time is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient, forming an actual surface point cloud; the surface three-dimensional point set of the virtual surgical planning model corresponding to the target bone anatomical site is obtained as the target point cloud; the initial alignment transformation matrix is ​​used as the initial transformation estimate of the iterative nearest point ICP algorithm to perform registration calculation on the actual surface point cloud and the target point set, and the optimized fine registration transformation matrix is ​​obtained to complete the fine registration; S3: Based on the fine registration transformation matrix, the virtual surgical planning model is superimposed and displayed in real time in the field of view of the mixed reality display device, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

[0025] To implement the above method, first deploy as follows: Figure 2 The mixed reality surgical navigation system shown is mainly composed of: an optically permeable mixed reality headset (such as HoloLens 2, as a client), a high-performance remote server (such as a dedicated workstation), a modular, personalized surgical guide (with QR code markings on its navigation components), and probes whose spatial position is tracked by an optical tracking system.

[0026] During the preoperative planning phase, the following tasks should be completed: Based on the patient's CT imaging data, a high-precision 3D virtual model of the patient's pelvis (especially the acetabulum) is reconstructed, and the ideal implantation position and orientation (i.e., reaming direction) of the acetabular cup is planned on this model. Based on the anatomical morphology of the acetabulum, a personalized patient-matching component is designed and 3D printed. Simultaneously, a standardized navigation registration component with QR code markings is assembled with the patient-matching component, and the spatial transformation relationship (matrix) between the two is precisely defined through calibration; this relationship is stored in the system. On the virtual pelvic model, a local anatomical reference frame is defined, and based on this reference frame, the planned reaming direction vector is calculated parametrically using the anatomical anteversion angle and the radiographic abduction angle.

[0027] Next, rapid registration is performed. The coordinate transformation relationships of each component in the entire rapid registration process are as follows: Figure 3 As shown, the coordinate systems involved include the world coordinate system. HoloLens coordinate system Probe coordinate system Guide plate coordinate system Patient coordinate system and the coordinate system of the virtual surgical planning model The key to the registration process lies in accurately determining the transformation relationship. It defines Compared to The position.

[0028] The first stage (step S1) is coarse registration based on a physical registration device, aiming to quickly obtain an initial alignment transformation matrix to align the virtual model to the real world. The initial alignment transformation matrix for aligning the virtual surgical planning model to the real-world coordinate system is calculated using the following formula: , in, The optical marker coordinate system of the physical registration device, determined during the preoperative planning phase. Relative to the coordinate system of the virtual surgical planning model The initial alignment transformation matrix; Intraoperative time The optical marker coordinate system obtained through tracking by the mixed reality display device The world coordinate system perceived by the mixed reality display device The real-time transformation matrix; Intraoperative time From the patient coordinate system to the optical marker coordinate system The real-time transformation matrix; Intraoperative time Obtained patient coordinate system Relative to the world coordinate system The real-time transformation matrix; The coordinate system of the virtual surgical planning model is calculated. Relative to the patient coordinate system The initial alignment transformation matrix, i.e., the coarse registration result.

[0029] In this THA embodiment, only the acetabular region can be exposed, and this region is a deeply concave hemispherical shape, making it difficult to accurately select predefined anatomical landmarks during the procedure. To address this issue, we designed a series of guiding components to perform coarse registration, such as... Figure 4 As shown. Through manual adjustment, a customized acetabular lateral wall matching component and a horseshoe fossa matching component are tightly attached to the corresponding area of ​​the pelvis, and Kirschner wires are inserted to fix the acetabular lateral wall matching component. The horseshoe fossa matching component is then removed, and a QR code marker (coordinate system) is installed on the acetabular lateral wall matching component. The grinding guide component is used. The relative position of the grinding guide component and the virtual pelvis / acetabulum has been accurately determined through preoperative planning, allowing for the establishment of a coordinate system. Compared to initial posture Specifically, the system obtains the coordinate system of the QR code marker in real time through visual recognition. The world coordinate system relative to the head-mounted display The transformation matrix is ​​then calculated. Simultaneously, the system retrieves the pre-labeled matrix from storage. Based on the formula mentioned above, the virtual model coordinate system is calculated. To the world coordinate system The initial alignment transformation matrix is ​​obtained. This process is completed automatically within seconds, achieving preliminary alignment (coarse registration) between the virtual model and the patient's anatomical structures.

[0030] Secondly, the second stage (step S2) involves fine registration based on point cloud and server computation. The goal is to refine the initial values ​​provided by the coarse registration. During the coarse registration process, the presence of osteophytes and soft tissue at the actual acetabulum may prevent the guide plate from being accurately placed in the intended position. Therefore, fine registration is necessary to achieve more precise alignment of the acetabulum. The Iterative Closest Point (ICP) algorithm is used to achieve finer alignment between the two coordinate systems. Specifically, this includes: The set of surface vertices of the target region in the preoperative 3D model is used as the target point cloud, where the first vertex is... The coordinates of each point are represented as follows: ; Multiple source point clouds were acquired on the surface of the target anatomical structure corresponding to the patient using a positioning probe, among which the first... The coordinates of each point are calculated using the following formula: , in, To collect the first At that moment, For a moment From the patient coordinate system To the world coordinate system The inverse of the transformation matrix, For a moment From the probe coordinate system To the world coordinate system The transformation matrix, The first probe acquisition The coordinates of the points; The rigid transformation matrix between the source point cloud and the target point cloud is calculated using the Iterative Closest Point (ICP) algorithm. Then the coordinate system of the virtual surgical planning model To the patient coordinate system The final fine registration transformation matrix is: ; To achieve real-time dynamic overlay of the simulated surgical planning model and the actual anatomical structure, the system calculates the coordinate system of the virtual surgical planning model in real time. To the coordinate system of mixed reality display devices Transformation matrix : , in, The coordinates obtained in real time by the augmented reality device through the SLAM self-localization system from the coordinate system of the mixed reality display device. To the world coordinate system The transformation matrix, For a moment From the patient coordinate system To the world coordinate system The transformation matrix.

[0031] The computation task of the Iterative Closest Point (ICP) algorithm in the fine registration is performed by a remote computing node that is communicatively connected to the mixed reality display device.

[0032] In this embodiment, the first step is point cloud acquisition: the doctor uses an optically tracked probe to continuously slide across the exposed surface of the acetabulum (especially characteristic areas such as the horseshoe fossa), acquiring approximately 50-800 three-dimensional spatial points to form an actual surface point cloud. Next, the headset client packages and sends the acquired data, along with the initial matrix obtained from coarse registration, to a remote server via a wireless network (using the HTTP / WebSocket protocol). This step offloads computationally intensive tasks from the headset, which has limited computing power.

[0033] After receiving the data, the server executes the ICP algorithm flow described in the pseudocode: It extracts the surface point set of the acetabular region from the virtual model as the target point cloud. Using the initial alignment transformation matrix as the initial transformation, it performs ICP calculations on both the source and target point clouds. The algorithm iteratively executes the steps of "finding the nearest point correspondence - calculating the optimal rigid transformation - updating the transformation" until the transformation update amount is less than a threshold (e.g., 1e-5) or the maximum number of iterations (e.g., 100 times) is reached. Finally, the server returns the final accurate transformation matrix and error data to the head-mounted display client via the network. The entire "send-calculate-return" process is optimized and controlled to within 10 seconds, far faster than the time required for equivalent calculations on the head-mounted display itself (approximately 3 minutes).

[0034] The third stage (step S3) involves virtual-real fusion and surgical navigation, utilizing the precise registration results to achieve real-time, accurate intraoperative guidance. In step S3, such as... Figure 5 As shown, the generation of the guidance instruction specifically includes: Establish a local anatomical reference system corresponding to the virtual surgical planning model The reference frame It consists of three orthogonal unit basis vectors. , and definition; Using the anatomical forward tilt angle located in the coordinate plane and imaging abduction angle Parameterized definition of the wear direction vector The calculation formula is: , in, The value of is 1 or -1, which is a sign parameter used to distinguish the spatial orientation of the target anatomical structure; Based on the precise registration transformation matrix, the wear direction vector is... Transform to real-world coordinate system Below, the actual guiding direction used for navigation display is obtained. : , in, This represents the overall transformation function.

[0035] In addition, the guidance instructions also include dynamic adjustment of guidance parameters during the procedure, including: During navigation, the system receives adjustment commands input by the user, including adjustments to the tilt angle. and abduction angle fine-tuning amount and ; The world coordinate system is updated in real time based on the aforementioned fine-tuning amount. The friction direction vector below The updated formula is: ; The virtual abrasion direction guide displayed on the mixed reality display device is recalculated and updated based on the updated direction vector.

[0036] In this embodiment, after receiving the final accurate transformation matrix, the head-mounted display client uses it as the final registration matrix. Based on this matrix, the system dynamically and stably overlays and renders the virtual pelvic model onto the corresponding position of the patient's real pelvis in the doctor's field of vision, achieving virtual-real fusion. Based on the pre-operatively planned grinding direction vector and the head-mounted display pose obtained through real-time tracking, the system calculates the direction of the navigation guide line to be displayed in the current field of vision in real time through a coordinate transformation chain (see the corresponding formula explanation section above), and renders it as a highlighted virtual guide rod at the center of the acetabulum. If the doctor needs to adjust the implantation angle during surgery according to the actual situation, they can input the fine adjustment amount of the anteversion or abduction angle via voice or gesture commands. The system will update the direction vector in real time according to the above update formula and immediately refresh the display of the navigation guide line. This function solves the inherent limitation of traditional physical guide plates being unable to be adjusted during surgery. Furthermore, and In practice, the values ​​are discrete. The system will display the values ​​of the anatomical anteversion angle and the radiographic abduction angle in real time after fine-tuning, to facilitate the surgeon's intraoperative operations.

[0037] To verify the effectiveness of the method of this invention, a comparative experiment with traditional registration methods was conducted. The experiment was performed in a simulated surgical environment, using the same pelvic model and hardware platform. The experimental results are as follows:

[0038] As can be seen, in terms of efficiency, this invention provides high-quality initial values ​​through coarse registration and offloads server computation, reducing the total registration time from the traditional 10 minutes to less than 2 minutes. Fine registration calculation takes only about 10 seconds, improving efficiency by over 80% and fully meeting clinical real-time requirements. In terms of accuracy, this invention achieves sub-millimeter level high accuracy (<2mm) comparable to traditional ICP algorithms. Furthermore, due to the introduction of coarse registration, the success rate and stability are significantly higher than the pure ICP method, which is sensitive to initial values. In terms of robustness, this invention avoids the difficult and error-prone manual marking within narrow acetabulums, overcoming the inherent limitations of traditional marker methods in deep surgical fields, and thus exhibiting greater adaptability.

[0039] In summary, the above methods fully demonstrate how the two-stage rapid registration method of the present invention, through the synergistic mechanism of physical coarse registration as the foundation and algorithmic fine registration as the enhancement, achieves an excellent balance between speed and accuracy under the constraints of real surgery, providing a practical solution for the clinical application of mixed reality surgical navigation.

[0040] Based on this, the present invention provides a rapid registration system for mixed reality navigation in orthopedic surgery, comprising: The coarse registration module is used to set optically recognizable marks on a physical registration device that is pre-fixed to the target anatomical site of the patient's bones. The marks are then identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. The fine registration module is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient using a probe whose spatial position is tracked in real time after the coarse registration is completed, forming an actual surface point cloud; to obtain the surface three-dimensional point set of the virtual surgical planning model corresponding to the target bone anatomical site as the target point cloud; and to perform registration calculation on the actual surface point cloud and the target point set using the initial alignment transformation matrix as the initial transformation estimate of the iterative nearest point ICP algorithm, so as to obtain the optimized fine registration transformation matrix and complete the fine registration. The fusion navigation module is used to overlay and display the virtual surgical planning model in real time on the field of view of the mixed reality display device according to the fine registration transformation matrix, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

[0041] It should be noted that the steps in the rapid registration method for mixed reality navigation in orthopedic surgery provided in this embodiment can be implemented based on the corresponding modules in the rapid registration system for mixed reality navigation in orthopedic surgery. Those skilled in the art can refer to the technical solution of the system to implement the steps of the method. That is, the embodiments in the system can be understood as preferred examples of implementing the method, and will not be elaborated here.

[0042] Besides implementing the system and its various devices provided by this invention in purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the system and its various devices of this invention appear as logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices provided by this invention can be considered as a hardware component, and the devices included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0043] Finally, it should be noted that the above description is only a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be pointed out that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

[0044] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A fast registration method for orthopedic surgery mixed reality navigation, characterized in that, Includes the following steps: S1: An optically recognizable mark is set on a physical registration device that is pre-fixed to the target anatomical site of the patient's bone. The mark is identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. S2: After completing the coarse registration, a probe whose spatial position is tracked in real time is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient to form an actual surface point cloud. Obtain the surface three-dimensional point set of the virtual surgical planning model corresponding to the target skeletal anatomical location, as the target point cloud; Using the initial alignment transformation matrix as the initial transformation estimate for the iterative nearest point ICP algorithm, the actual surface point cloud and the target point set are registered to obtain the optimized fine registration transformation matrix, thus completing the fine registration. S3: Based on the fine registration transformation matrix, the virtual surgical planning model is superimposed and displayed in real time in the field of view of the mixed reality display device, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

2. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 1, characterized in that, In step S1, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real-world coordinate system is calculated using the following formula: ， in, The optical marker coordinate system of the physical registration device, determined during the preoperative planning phase. Relative to the coordinate system of the virtual surgical planning model The initial alignment transformation matrix; Intraoperative time The optical marker coordinate system obtained through tracking by the mixed reality display device The world coordinate system perceived by the mixed reality display device The real-time transformation matrix; Intraoperative time From the patient coordinate system to the optical marker coordinate system The real-time transformation matrix; Intraoperative time Obtained patient coordinate system Relative to the world coordinate system The real-time transformation matrix; The coordinate system of the virtual surgical planning model is calculated. Relative to the patient coordinate system The initial alignment transformation matrix, i.e., the coarse registration result.

3. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 2, characterized in that, In step S2, based on the coarse registration result, the fine registration is performed using the Iterative Closest Point (ICP) algorithm, specifically including: The set of surface vertices of the target region in the preoperative 3D model is used as the target point cloud, where the first vertex is... The coordinates of each point are represented as follows: ; Multiple source point clouds were acquired on the surface of the target anatomical structure corresponding to the patient using a positioning probe, among which the first... The coordinates of each point are calculated using the following formula: ， in, To collect the first At that moment, For a moment From the patient coordinate system To the world coordinate system The inverse of the transformation matrix, For a moment From the probe coordinate system To the world coordinate system The transformation matrix, The first probe acquisition The coordinates of the points; The rigid transformation matrix between the source point cloud and the target point cloud is calculated using the Iterative Closest Point (ICP) algorithm. Then the coordinate system of the virtual surgical planning model To the patient coordinate system The final fine registration transformation matrix is: ； To achieve real-time dynamic overlay of the virtual surgical planning model and the actual anatomical structure, the system calculates the coordinate system of the virtual surgical planning model in real time. To the coordinate system of mixed reality display devices Transformation matrix : ， in, The coordinates obtained in real time by the augmented reality device through the SLAM self-localization system from the coordinate system of the mixed reality display device. To the world coordinate system The transformation matrix, For a moment From the patient coordinate system To the world coordinate system The transformation matrix.

4. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 3, characterized in that, In step S3, the generation of the guidance instruction specifically includes: Establish a local anatomical reference system corresponding to the virtual surgical planning model The reference frame It consists of three orthogonal unit basis vectors. , and definition; Using the anatomical forward tilt angle located in the coordinate plane and imaging abduction angle Parameterized definition of the wear direction vector The calculation formula is: ， in, The value of is 1 or -1, which is a sign parameter used to distinguish the spatial orientation of the target anatomical structure; Based on the precise registration transformation matrix, the wear direction vector is... Transform to real-world coordinate system Below, the actual guiding direction used for navigation display is obtained. : ， in, This represents the overall transformation function.

5. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 4, characterized in that, The guidance instructions also include dynamic adjustment of guidance parameters during the procedure, including: During navigation, the system receives adjustment commands input by the user, including adjustments to the tilt angle. and abduction angle fine-tuning amount and ; The world coordinate system is updated in real time based on the aforementioned fine-tuning amount. The friction direction vector below The updated formula is: ； The virtual abrasion direction guide displayed on the mixed reality display device is recalculated and updated based on the updated direction vector.

6. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 3, characterized in that, In step S2, the computation task of the Iterative Closest Point (ICP) algorithm in the fine registration is performed by a remote computing node that is communicatively connected to the mixed reality display device.

7. The rapid registration method for mixed reality navigation in orthopedic surgery according to claim 2, characterized in that, The physical registration device is a modular, personalized surgical guide, comprising a fitting component individually matched to the patient's skeletal anatomy and a navigation component with the optical markers. The fitting component and the navigation component are detachably connected, and the initial alignment transformation relationship... The connection between the two and the virtual surgical planning model are determined before the operation.

8. A rapid registration system for orthopedic surgery using mixed reality navigation, characterized in that, include: The coarse registration module is used to set optically recognizable marks on a physical registration device that is pre-fixed to the target anatomical site of the patient's bones. The marks are then identified by a mixed reality display device to obtain the first pose of the physical registration device in the real space coordinate system. Based on the fixed pose relationship between the physical registration device and the virtual surgical planning model that has been pre-stored in the preoperative planning, the initial alignment transformation matrix for aligning the virtual surgical planning model to the real space coordinate system is calculated to complete the coarse registration. The fine registration module is used to continuously collect a predetermined number of three-dimensional spatial points on the actual bone surface of the target bone anatomical site of the patient after the coarse registration is completed, using a probe whose spatial position is tracked in real time, to form an actual surface point cloud. Obtain the surface three-dimensional point set of the virtual surgical planning model corresponding to the target skeletal anatomical location, as the target point cloud; Using the initial alignment transformation matrix as the initial transformation estimate for the iterative nearest point ICP algorithm, the actual surface point cloud and the target point set are registered to obtain the optimized fine registration transformation matrix, thus completing the fine registration. The fusion navigation module is used to overlay and display the virtual surgical planning model in real time on the field of view of the mixed reality display device according to the fine registration transformation matrix, so that it is precisely aligned with the patient's real anatomical structure in space; based on the fine registration transformation matrix and the pre-stored preoperative planning parameters, virtual instrument operation guidance instructions are generated and displayed in the field of view of the mixed reality display device, and the guidance instructions include at least a virtual abrasion direction vector representing the expected operation direction.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the rapid registration method for mixed reality navigation in orthopedic surgery as described in any one of claims 1-7.

10. An electronic device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the rapid registration method for mixed reality navigation in orthopedic surgery as described in any one of claims 1-7.

Images