A positioning device for registering tissue and blood vessels in conjunction with a head-mounted holographic display
Patent Information
- Application Number
- CN202520964152.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-16
- Publication Date
- 2026-09-25
- Estimated Expiration
- 2035-05-16
AI Technical Summary
[0004]本实用新型的目的是为了解决上述背景技术中提出的定位标志物无法满足低成本、高精度、高效率要求的问题,而提出的一种与头戴式全息显示器结合配准组织和血管的定位装置
[0012]本实用新型中装置结构简单,运用可靠,克服了原有虚拟定位标志物需使用繁琐且昂贵的外部辅助设备。同时,可以印制真实二维码于人体表面,该二维码可被消毒至无菌,可在术中实时导航使用,同时印制无创,克服了现有实体定位标志物需在术前被去除以消毒至无菌、只可在术前进行配准、定位标志物有创植入等一系列缺点,可以应用在人体任何区域且无创。
Smart Images

Figure CN224792337U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of medical device technology, and in particular to a positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display. Background Technology
[0002] Current methods for registering human tissue with its 3D model primarily involve virtual marker registration and physical marker registration. Virtual marker registration often requires equipment such as optical positioning systems, which are expensive and can be overwhelming for the surgeon during surgery, resulting in high costs. Physical markers include skeletal landmarks and electrode pads. Skeletal landmarks are mainly used for locations with fixed anatomical landmarks, such as the eyes, ears, nose, and throat. However, in smooth, flat human tissues without clear fixed anatomical landmarks, localization is impossible. Electrode pads, as physical markers, are primarily made of metal, which produces artifacts after image scanning, leading to higher registration errors.
[0003] Therefore, existing positioning markers cannot meet the requirements of low cost, high precision, and high efficiency. Utility Model Content
[0004] The purpose of this invention is to address the problem that the positioning markers mentioned in the background art cannot meet the requirements of low cost, high precision, and high efficiency, and to propose a positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display.
[0005] To achieve the above objectives, the present invention adopts the following technical solution:
[0006] A positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display includes a sticker layer and a positioning layer. The sticker layer is adhered to the upper surface of the positioning layer. A QR code layer is printed on the side of the positioning layer near the sticker layer, and a “┏”-shaped imaging layer is adhered to the side of the positioning layer away from the QR code layer.
[0007] Preferably, the sticker layer is a polyurethane film layer.
[0008] Preferably, the positioning layer is a PET film layer.
[0009] Preferably, the side lengths of the “┏”-shaped developing layers are equal and are half the side length of the QR code layer.
[0010] Preferably, the “┏”-shaped developing layer is a PEEK material layer.
[0011] Compared with the prior art, the present invention has the following beneficial effects:
[0012] The device in this invention has a simple structure and reliable operation, overcoming the need for cumbersome and expensive external auxiliary equipment required by traditional virtual positioning markers. Simultaneously, a real QR code can be printed on the human body surface. This QR code can be sterilized and used for real-time navigation during surgery. Furthermore, the printing process is non-invasive, overcoming a series of drawbacks of existing physical positioning markers, such as the need for pre-operative removal for sterilization, the limitation to pre-operative registration, and the invasive implantation of positioning markers. This device can be applied to any area of the human body without causing injury. Attached Figure Description
[0013] Figure 1 This is a schematic diagram of the structure of a positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display, as proposed in this utility model.
[0014] Figure 2 This is a schematic diagram of the "┏"-shaped imaging layer in a positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display, as proposed in this utility model.
[0015] In the diagram: 1. Sticker layer, 2. Positioning layer, 3. QR code layer, 4. "┏" shaped developing layer. Detailed Implementation
[0016] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present utility model. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments.
[0017] Reference Figure 1-2 A positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display includes a sticker layer 1 and a positioning layer 2, wherein the sticker layer 1 is adhered to the upper surface of the positioning layer 2.
[0018] In this embodiment, the sticker layer 1 is a polyurethane film layer, the positioning layer 2 is a PET film layer, and the side of the positioning layer 2 closest to the sticker layer 1 is printed with a QR code layer 3.
[0019] In this embodiment, a “┏”-shaped developing layer 4 is bonded to the side of the positioning layer 2 away from the QR code layer 3. The side length of the “┏”-shaped developing layer 4 is equal to that of the side length of the QR code layer 3, and the “┏”-shaped developing layer 4 is a PEEK material layer.
[0020] In this embodiment, if it is necessary to understand the bones, blood vessels, etc. of a specific location on the patient, the device can be pasted on the target area for imaging examination. Then, the obtained image data is imported into the software workstation for three-dimensional reconstruction. According to the position of the three-dimensional model of the "┏" shape, a virtual QR code that is exactly the same as the one printed on the human body is placed in contact with the "┏" shape. At this time, the position of the virtual QR code in the three-dimensional model is exactly the same as the position of the actual QR code on the human body. The three-dimensional model of the "┏" shape is removed. The positioning device pasted on the human body is removed before or during the operation. At this time, the device has printed the real QR code on the surface of the human skin tissue. The registration of the human body and its three-dimensional model can be achieved by registering the real and virtual QR codes.
[0021] In this embodiment, the device has a simple structure and reliable operation, overcoming the need for cumbersome and expensive external auxiliary equipment required by traditional virtual positioning markers. Simultaneously, a real QR code can be printed on the human body surface. This QR code can be sterilized and used for real-time navigation during surgery. Furthermore, the printing process is non-invasive, overcoming a series of drawbacks of existing physical positioning markers, such as the need for pre-operative removal for sterilization, the limitation of pre-operative registration, and the invasive implantation of positioning markers. It can be applied to any area of the human body without causing injury.
[0022] Example 2:
[0023] In this embodiment, upon patient admission, the sticker film of the device is peeled off, and the device is affixed to the target area. The patient undergoes imaging examinations to obtain DICOM data. This data is then imported into a software workstation for 3D reconstruction, creating a 3D model of the human skeleton, blood vessels, soft tissue, and the "┏" shaped material on the back of the positioning device. This 3D model is then processed in the software workstation. Using the position of the "┏" shaped material as a positioning reference, an electronic virtual QR code, identical to the one printed on the human body, is inserted at this location. The positional relationship between this virtual QR code and the 3D model of the "┏" shaped material is completely consistent with the position of the actual QR code and the "┏" shaped material in the positioning device. Simultaneously, this virtual QR code is aligned with the 3D model of the human skin. The 3D model of the "┏" shaped material is then removed in the software workstation. At this point, the position of the virtual QR code on the human 3D model is exactly the same as the position of the actual QR code on the human body.
[0024] In this embodiment, an artificial intelligence algorithm is developed within the application of a head-mounted holographic display. This algorithm includes the YOLO algorithm and the Kalman filter algorithm. The YOLO algorithm automatically registers a virtual QR code in a 3D model with a real QR code printed on the human body. After registration, the human body and its 3D model completely overlap, achieving 3D visualization of human tissue. The Kalman filter algorithm tracks and re-matches the real QR code printed on the human body, ensuring that even when the observer's head moves while wearing the head-mounted holographic display, the virtual and real QR codes remain registered in real time. After registration, the 3D model can be overlaid on the human body, achieving 3D visualization of human tissue. The specific process is as follows:
[0025] ① YOLO Algorithm for Registration of Real and Virtual QR Codes: In cross-modal registration of real QR codes on the patient's body surface and virtual QR codes in a 3D model, the system employs an improved YOLOv8 architecture to construct a multi-scale feature fusion network. The network input layer receives 1080P real-time image streams captured by an intraoperative RGB camera and projection images of virtual QR codes generated by CTA 3D reconstruction. Texture and spatial geometric features at different levels are extracted through a feature pyramid network. To address common intraoperative non-uniform lighting interference, an adaptive histogram equalization module is introduced in the preprocessing stage to dynamically enhance the contrast threshold between the QR code edges and the background. The detection head uses deformable convolutional kernels instead of traditional convolutional layers, enabling the network to automatically adapt to local deformations of the QR code caused by respiratory movements or instrument compression. The feature decoder synchronously outputs the 2D pixel coordinates of the QR code vertices and their corresponding 3D spatial projection matrices. During training, the system uses a synthetic data generation engine to simulate complex scenarios such as changes in operating room lighting, blood contamination, and instrument occlusion. By randomly generating segmentation masks covering 30%-50% of the original QR code surface, the system forces the network to learn local feature inference capabilities, ultimately achieving an intraoperative detection success rate of over 95% and sub-pixel-level positioning accuracy. Upon detecting a QR code in the real environment, the system calculates the rigid transformation matrix between it and the virtual model, and eliminates registration errors introduced by soft tissue deformation through iterative nearest-point optimization, completing the initial spatial alignment of the augmented reality scene.
[0026] ② The Kalman filter algorithm completes the tracking and re-matching during the movement of the head-mounted holographic display: The nine-axis inertial measurement unit built into the head-mounted display continuously outputs the observer's pose change data at a frequency of 200Hz. The system uses this as the observation input of the Kalman filter to construct a twelve-dimensional state vector, which includes the virtual QR code's position coordinates (x, y, z) and Euler angle attitude (α, β, γ) in three-dimensional space, as well as its corresponding linear and angular velocity components. In the filter prediction stage, the pose prior estimate of the virtual QR code is calculated based on the head-mounted display's kinematic model. The process noise covariance matrix adopts a dynamic adjustment mechanism: when the IMU accelerometer detects rapid movement of the observer (linear acceleration > 2m / s²), the dynamic adjustment mechanism is used.2 When the gyroscope detects slow rotation (angular velocity < 5° / s), the noise weight of the position component is automatically increased to prioritize the visual observation data. When the gyroscope detects slow rotation (angular velocity < 5° / s), the attitude angle update weight is increased to suppress the cumulative drift error of the vision system. During the measurement update phase, the coordinates of the QR code vertex detected in real time by the YOLO algorithm are bidirectionally matched with the reprojected coordinates of the virtual model. When the distance between the two in the image plane exceeds a preset threshold, the pose correction module based on Lie group optimization is triggered, and six-degree-of-freedom parameter iterative optimization is performed in the manifold space to ensure that the dynamic registration accuracy can be maintained below 0.8mm under conditions of violent movement or brief occlusion. Finally, the system integrates the prediction results of the Kalman filter and the visual correction amount, and generates a smooth pose transformation sequence through quaternion interpolation to drive the head-mounted display to achieve lag-free augmented reality rendering.
[0027] In this embodiment, production practice has demonstrated that the invention can effectively improve the efficiency and accuracy of registration between the human body and its three-dimensional model, and can achieve complete sterilization of the surgery to a sterile state, and can provide real-time navigation during the operation.
[0028] The above description is only a preferred embodiment of the present utility model, but the protection scope of the present utility model is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present utility model, based on the technical solution and the inventive concept of the present utility model, should be included within the protection scope of the present utility model.
Claims
1. A positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display, comprising a sticker layer (1) and a positioning layer (2), characterized in that: The sticker layer (1) is adhered to the upper surface of the positioning layer (2). A QR code layer (3) is printed on the side of the positioning layer (2) close to the sticker layer (1), and a "┏" shaped developing layer (4) is adhered on the side of the positioning layer (2) away from the QR code layer (3).
2. The positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display according to claim 1, characterized in that: The sticker layer (1) is a polyurethane film layer.
3. The positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display according to claim 1, characterized in that: The positioning layer (2) is a PET film layer.
4. The positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display according to claim 1, characterized in that: The side lengths of the “┏”-shaped developing layer (4) are equal and are half the side lengths of the QR code layer (3).
5. The positioning device for registering tissues and blood vessels in conjunction with a head-mounted holographic display according to claim 1, characterized in that: The “┏” shaped developing layer (4) is a PEEK material layer.