Pulmonary nodule auxiliary positioning method, system, device, storage medium and product
By constructing a preoperative virtual lung model and generating a dynamic 3D model by combining thoracoscopic images, augmented reality technology is used to achieve precise localization of pulmonary nodules, solving the localization problem in existing technologies, improving accuracy and reducing patient suffering.
Patent Information
- Application Number
- CN202610630750.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-09
- Publication Date
- 2026-07-03
AI Technical Summary
In current technologies, it is difficult to accurately locate pulmonary nodules, especially in thoracoscopic surgery where effective palpation is impossible. Furthermore, CT-guided localization with a localization needle is painful and has a high failure rate, particularly in certain areas such as the lung apex and near major blood vessels.
By constructing a preoperative virtual lung model of the patient and combining it with lesion images acquired by thoracoscopic imaging, an intraoperative dynamic 3D model is generated. Combined with real-time intraoperative images, the model helps predict the location of pulmonary nodules in the patient's solid lung, and uses augmented reality technology to achieve precise localization of pulmonary nodules.
It improves the accuracy of pulmonary nodule localization, reduces patient suffering and waiting time, lowers the localization failure rate, and avoids additional harm to patients.
Smart Images
Figure CN122320677A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of lung nodule localization technology, and in particular to a method, system, electronic device, computer storage medium, and computer program product for assisted localization of small lung nodules. Background Technology
[0002] In recent years, an increasing number of patients have sought medical attention after small pulmonary nodules were discovered during physical examinations. Domestic and international literature reports that the malignancy rate of solid pulmonary nodules (SPNs) is 20%–40%, while the malignancy rate of mixed ground-glass nodules (mGGOs) is even higher. Therefore, it is currently believed that high-risk GGOs in the lungs should be surgically removed at an early stage.
[0003] Thoracoscopic segmentectomy and wedge resection are currently the preferred methods widely recognized in the medical community for the definitive diagnosis and treatment of pulmonary nodules. Pulmonary nodule surgery often employs a minimally invasive thoracoscopic approach, offering advantages such as minimal trauma and rapid recovery. However, because pulmonary nodules located within the lung parenchyma are small, most are not visible on the lung surface and are difficult to locate by palpation during thoracoscopic surgery.
[0004] Currently, CT-guided preoperative hook-wire localization of pulmonary nodules requires the patient to be conscious during and after localization, which is quite painful for the patient. It also requires transfer from the CT room to the operating room, resulting in a long waiting time and increasing the risk of various complications. For certain areas, such as the apex of the lung and near major blood vessels, localization is difficult and has a high failure rate.
[0005] This paper proposes a method, system, electronic device, computer storage medium, and computer program product for assisted localization of pulmonary nodules. Summary of the Invention
[0006] This specification provides a method, system, electronic device, computer storage medium, and computer program product for assisting in the localization of pulmonary nodules, in order to improve the accuracy of pulmonary nodule localization.
[0007] The method for assisted localization of pulmonary nodules provided in this application adopts the following technical solution, including: Construct a preoperative virtual model of the patient's lungs; Images of the lesion were acquired via thoracoscopy; Based on the preoperative lung virtual model and the lesion image, an intraoperative dynamic 3D model is generated; Real-time intraoperative images are acquired using the thoracoscopic technique; By combining the intraoperative dynamic 3D model with the intraoperative real-time images, the location of pulmonary nodules in the patient's solid lung can be predicted.
[0008] Optionally, constructing the patient's preoperative virtual lung model includes: Obtain the patient's preoperative CT data; Based on the preoperative CT data, a three-dimensional reconstruction is performed to generate the preoperative lung virtual model; Virtual lung nodules are marked in the preoperative virtual lung model.
[0009] Optionally, generating an intraoperative dynamic 3D model of the lung lobe where the lesion is located based on the preoperative virtual lung model and the lesion image includes: Spatially align the preoperative lung virtual model with the lesion image; Based on the spatial alignment results and the patient's respiratory movements, an intraoperative dynamic 3D model of the lung lobe where the lesion is located is generated.
[0010] Optionally, spatially aligning the preoperative lung virtual model with the lesion image includes: Identify common features between the lesion image and the preoperative lung virtual model; The lesion image is aligned with the preoperative lung virtual model based on the common markers.
[0011] Optionally, the step of combining the intraoperative dynamic 3D model with the intraoperative real-time images to assist in predicting the location of pulmonary nodules in the patient's solid lung includes: Real-time images of the patient's actual lung region are obtained based on the intraoperative real-time images; The intraoperative dynamic 3D model is fused with the actual lung area using the real-time video feed. Based on the depth and angle information of the virtual lung nodule relative to the patient's body surface, the location of the lung nodule in the patient's physical lung is assisted.
[0012] Optional, also includes: Lung nodules are located in real time based on the activity of the patient's solid lungs.
[0013] The pulmonary nodule-assisted localization system provided in this application adopts the following technical solution, including: The preoperative diseased lung modeling module is used to construct a virtual model of the patient's preoperative lungs; Thoracoscopic video input module, used to acquire images of the lesion through thoracoscopy; The intraoperative lung modeling module is used to generate an intraoperative dynamic 3D model based on the preoperative virtual lung model and the lesion image. The video acquisition module is used to acquire real-time intraoperative images through the thoracoscope; The intelligent positioning module is used to combine the intraoperative dynamic 3D model with the intraoperative real-time images to help predict the location of pulmonary nodules in the patient's solid lung.
[0014] Optionally, the preoperative diseased lung modeling module includes: The data acquisition submodule is used to acquire the patient's preoperative CT data; The three-dimensional reconstruction submodule is used to perform three-dimensional reconstruction based on the preoperative CT data and generate the preoperative lung virtual model. The tagging submodule is used to tag virtual lung nodules in the preoperative lung virtual model.
[0015] Optionally, the intraoperative lung modeling module includes: The alignment submodule is used to spatially align the preoperative lung virtual model with the lesion image; The model building submodule is used to generate an intraoperative dynamic 3D model of the lung lobe where the lesion is located based on the spatial alignment results and the patient's respiratory movements.
[0016] Optionally, the alignment submodule includes: The marker lookup unit is used to find common markers between the lesion image and the preoperative lung virtual model; Alignment unit, used to align the lesion image with the preoperative lung virtual model based on the common marker.
[0017] Optionally, the intelligent positioning module includes: The image acquisition submodule is used to acquire real-time images of the patient's actual lung region based on the intraoperative real-time images; The overlay and fusion submodule is used to fuse the intraoperative dynamic 3D model with the actual lung region using the real-time image; The positioning submodule is used to assist in locating the position of the pulmonary nodule in the patient's physical lung based on the depth and angle information of the virtual pulmonary nodule relative to the patient's body surface.
[0018] Optional, also includes: The lung nodule real-time location tracking module is used to track and locate lung nodules in real time based on the activity of the patient's solid lungs.
[0019] This specification also provides a computer device, wherein the computer device includes: Processor; and, A memory that stores computer-executable instructions, which, when executed, cause the processor to perform any of the methods described above.
[0020] This specification also provides a computer-readable storage medium that stores one or more programs / instructions that, when executed by a processor, implement any of the methods described above.
[0021] This specification also provides a computer program product, wherein the computer program product includes: a computer program / instruction, which, when executed by a processor, implements any of the methods described above.
[0022] In this invention, a preoperative virtual lung model of the patient is constructed; images of the lesion are acquired via thoracoscopy; an intraoperative dynamic 3D model is generated based on the preoperative virtual lung model and the lesion images; real-time intraoperative images are acquired via thoracoscopy; and the intraoperative dynamic 3D model and the real-time intraoperative images are combined to assist in predicting the location of pulmonary nodules in the patient's solid lung, thereby improving the accuracy of pulmonary nodule localization. Attached Figure Description
[0023] Figure 1 A schematic diagram illustrating the principle of a lung nodule-assisted localization method provided in the embodiments of this specification; Figure 2 This is a schematic diagram of a lung nodule-assisted localization system provided in an embodiment of this specification; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification; Figure 4 This is a schematic diagram of a computer-readable storage medium provided for an embodiment of this specification. Detailed Implementation
[0024] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.
[0025] Exemplary embodiments of the invention will now be described more fully with reference to the accompanying drawings. While conforming to the inventive concept, the features, structures, characteristics, or other details described in a particular embodiment may be combined in one or more other embodiments in a suitable manner.
[0026] The terms “and / or” or “and / or” include all combinations of any one or more of the listed items.
[0027] If the technical solution of this application involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.
[0028] Figure 1 A schematic diagram illustrating the principle of a lung nodule-assisted localization method provided in the embodiments of this specification includes: S1 constructs a virtual model of the patient's preoperative lungs; S2 acquires images of the lesion via thoracoscopy; S3 generates an intraoperative dynamic 3D model based on the preoperative lung virtual model and the lesion image; S4 acquires real-time intraoperative images via the thoracoscope; S5 combines the intraoperative dynamic 3D model with the intraoperative real-time images to help predict the location of pulmonary nodules on the surface of the patient's solid lung.
[0029] As mentioned earlier, accurately determining the location of small pulmonary nodules during wedge resection surgery is a pressing clinical problem. To address the shortcomings of existing technologies, this invention provides a method for assisted localization of small pulmonary nodules, solving the technical problem of inaccurate localization of the nodule region, and is used for automatic assisted navigation and localization of small pulmonary nodules. Specifically: S1 constructs a virtual model of the patient's preoperative lungs; S11 Acquire the patient's preoperative CT data; Preoperative CT data is DICOM format image data acquired through CT scans. Preoperative CT data includes: patient information, scan parameters, and several two-dimensional tomographic images. The two-dimensional tomographic images (DICOM images) are pixel matrices, with each pixel value represented by a corresponding CT value (HU value).
[0030] S12 performs preprocessing on the preoperative CT data; S121 performs format conversion on the preoperative CT data; In one embodiment of this specification, preoperative CT data is read, and scanning parameters and several two-dimensional tomographic images are extracted; the scanning parameters include slice thickness.
[0031] Two-dimensional tomographic images with different scanning layer thicknesses are resampled to a uniform layer thickness to avoid geometric distortion during three-dimensional reconstruction caused by uneven layer thickness.
[0032] S122 performs noise removal on the preoperative CT data after format conversion; Considering that CT data is susceptible to electronic noise, causing pixel values to fluctuate randomly, in one embodiment of this specification, a filtering algorithm is used to smooth the two-dimensional tomographic image while preserving the edge details of lung tissue.
[0033] If a patient has a metal implant (such as a cardiac stent), deep learning-based artifact correction algorithms or iterative reconstruction techniques can be used to reduce the interference of metal artifacts on the lung parenchyma.
[0034] S123 normalizes the preoperative CT data after noise removal; Lung tissue appears as low density in CT data, with a HU value of approximately -1000 to -400, but the HU value range may vary between different scanning devices. Therefore, in one embodiment of this specification, the HU value of the lung parenchyma is normalized by linear transformation to improve the stability of subsequent segmentation algorithms by unifying the data scale.
[0035] S13 performs three-dimensional reconstruction based on the preoperative CT data to generate the preoperative lung virtual model; S131 performs tissue segmentation on the preoperative CT data based on a tissue segmentation model; The lung parenchyma is the main functional area of the lung and needs to be separated from the thoracic cavity tomographic images that include structures such as the chest wall, mediastinum, and heart.
[0036] In one embodiment of this specification, the tissue segmentation model is a threshold segmentation algorithm; specifically, taking advantage of the low density characteristics of the HU value of lung tissue, a threshold range is set (e.g., -1000 to -400 HU), pixels below the threshold range are marked as lung parenchyma, and pixels above the threshold range are marked as other tissues (such as chest wall, heart, etc.). In another embodiment of this specification, the tissue segmentation model is a region growth model; specifically, using the hilar region (pulmonary artery / vein inlet) as a seed point, it expands outwards based on pixel grayscale similarity, gradually growing until it covers the entire lung parenchyma. The remaining locations are marked as other tissues.
[0037] In another embodiment of this specification, the tissue segmentation model is a deep learning segmentation model. Specifically, for complex anatomical structures such as lung fissures and interlobar vessels, a pre-trained deep learning model is used for end-to-end segmentation. Through training on a large amount of labeled lung CT data, the model learns the differences in edge and texture features between the lung parenchyma and surrounding tissues, thereby distinguishing the lung parenchyma from other tissues. The pre-trained deep learning model includes, but is not limited to, 3D U-Net and nnU-Net.
[0038] S132 performs structural localization of the lung parenchyma in the preoperative CT data based on the structural segmentation model; Precise localization of lung anatomy requires further segmentation of lung lobes, lung segments, bronchi, and blood vessels as key references for locating lung nodules.
[0039] S132-1 identifies lung lobes and lung segments in the lung parenchyma of the preoperative CT data; In one embodiment of this specification, the structural segmentation model is a multi-task learning model. By using a multi-task learning model, the accuracy of lung lobe segmentation masks and interlobar fissure segmentation results are output simultaneously, thereby improving the accuracy of lobe segment boundaries.
[0040] S132-2 identifies bronchi in the lung parenchyma of the preoperative CT data; In one embodiment of this specification, the structural segmentation model is a centerline tracing method. This method identifies bronchi in the lung parenchyma from preoperative CT data. Specifically, tubular structures are extracted using a Frangi filter, and then morphological refinement is applied to obtain the bronchial centerline, thereby achieving bronchial identification. To avoid segmented bronchial display caused by thick CT slices, a deep learning model can be incorporated to optimize branch breakpoint repair.
[0041] S132-3 identifies blood vessels in the lung parenchyma of the preoperative CT data; Considering that small blood vessels (diameter <2mm) are easily masked by noise in low-dose CT.
[0042] In one embodiment of this specification, the structural segmentation model is a multi-scale feature extraction model. Multi-scale feature extraction is used. In another embodiment of this specification, contrast enhancement is used to improve the contrast between blood vessels and surrounding tissues.
[0043] S133 combines tissue segmentation results with structural segmentation results to perform virtual lung surface reconstruction and generate a preoperative virtual lung model. In one embodiment of this specification, preoperative CT data is traversed, and based on lung parenchyma markers, it is determined whether the data passes through a structural surface in each voxel; isosurfaces are extracted and triangular meshes are generated. Specifically, the transition of pixel values within a voxel from 0 to 1 or from 1 to 0 determines whether the data passes through a structural surface.
[0044] S14 marks virtual lung nodules in the preoperative virtual lung model.
[0045] Regions with abnormal HU values within the lung parenchyma are marked as virtual lung nodules. This invention achieves precise labeling by integrating the spatial information of pulmonary nodules into a preoperative virtual lung model. The spatial information of pulmonary nodules includes, but is not limited to: location / central coordinates, size (such as maximum diameter, volume), and distance from surrounding tissues (pleura, blood vessels).
[0046] This invention utilizes individualized virtual lung reconstruction based on CT images. After a CT scan of the patient, the lung CT scan images are reconstructed in 3D to mark the location of lung nodules, which are then used as a reference for lung nodule localization.
[0047] S2 acquires images of the lesion via thoracoscopy; During the procedure, details of the patient's lesion area (such as the surface morphology of lung tissue and the surface projection of nodule location) are captured in real time using a thoracoscopic lens and used as lesion images for lung nodule localization.
[0048] S3 generates an intraoperative dynamic 3D model based on the preoperative lung virtual model and the lesion image; S31 Spatially aligns the preoperative lung virtual model with the lesion image; S311 Identify common features between the lesion image and the preoperative lung virtual model; Common landmarks include, but are not limited to: the edge of the lung lobe, the bifurcation point of the bronchus, and the course of blood vessels.
[0049] S312 aligns the lesion image with the preoperative lung virtual model based on the common marker.
[0050] Based on common markers, spatial coordinate alignment of the "preoperative lung virtual model - patient's actual lung" is achieved, and the projection position of the preoperative lung virtual model on the patient's body surface is determined.
[0051] This invention achieves the effect of "projecting the image onto the body surface" by spatially calibrating the image pixel coordinates with the patient's intraoperative body surface coordinate system by identifying common landmarks in the thoracoscopic field of view.
[0052] Based on the spatial alignment results and the patient's respiratory movements, S32 generates an intraoperative dynamic 3D model of the lung lobe where the lesion is located.
[0053] Considering that respiratory motion causes periodic fluctuations in lung tissue, directly affecting intraoperative positioning accuracy, in one embodiment of this specification, by combining real-time lung tissue motion in intraoperative thoracoscopic images (e.g., lobe undulation caused by respiration), and through time series analysis, the temporal changes in intraoperative images (e.g., lobe displacement and deformation between consecutive frames) are captured to generate an intraoperative dynamic 3D model of the lung lobe where the lesion is located. This allows for the overlay of real-time intraoperative deformation information onto the preoperative virtual lung model, compensating for the influence of respiratory motion.
[0054] Specifically, by optically tracking chest wall movement in the thoracoscopic field of view and simultaneously marking the respiratory cycle, the temporal synchronization of the dynamic model is further improved.
[0055] S4 acquires real-time intraoperative images via the thoracoscope; Since thoracoscopic images need to be continuously acquired during the operation, dynamic images of the patient's chest cavity are acquired in real time through the thoracoscopic lens to facilitate data processing.
[0056] S5 combines the intraoperative dynamic 3D model with the intraoperative real-time images to help predict the location of pulmonary nodules on the surface of the patient's solid lung.
[0057] S51 acquires a real-time image of the patient's actual lung region based on the intraoperative real-time images; Thoracoscopic imaging is essentially a "window" into the patient's body surface—the lens enters the chest cavity, and the real-time image it captures can be seen as a "local projection" of the patient's chest wall surface. Therefore, the real-time image is the image projected directly onto the patient's body surface from the thoracoscopy.
[0058] S52 uses the real-time image to fuse the intraoperative dynamic 3D model with the actual lung area; Using augmented reality (AR) devices, a dynamic 3D model of the lung lobe where the lesion is located is superimposed on the real-time image of the patient's actual lung area, so that the virtual lung nodule and the projection position of the solid lung surface completely overlap.
[0059] S53 uses the depth and angle information of the virtual lung nodule relative to the patient's body surface to help predict the position of the lung nodule in the patient's solid lung.
[0060] At this point, the user's field of vision displays a visual overlay of "virtual lung nodules and solid lung." By calculating the depth (i.e., the depth of the nodule's embedment within the pleural cavity) and angle (such as the angle with the pleura and ribs) of the virtual lung nodule relative to the patient's body surface, the location of the small lung nodule on the surface of the solid lung is predicted. The user is the doctor who locates and treats the small lung nodule on the patient.
[0061] S531 calculates the depth information of the virtual lung nodule relative to the patient's body surface, as the depth information of the lung nodule; In one embodiment of this specification, the coordinates of the virtual nodule are converted into coordinates in the intraoperative body surface coordinate system, and the Euclidean distance between the nodule and the chest wall is directly calculated as the depth information of the pulmonary nodule.
[0062] S532 calculates the angle information of the virtual lung nodule relative to the patient's body surface, as the angle information of the lung nodule; In one embodiment of this specification, the surface equation of the pleura and the orientation of the ribs are extracted from the intraoperative dynamic 3D model to establish the geometric parameters of the anatomical structure, including the normal vector and tangent direction. Taking the center of the virtual nodule as the vertex, the angles between it and the pleural normal vector and the rib tangent are calculated using the aforementioned geometric parameters to obtain the angular relationship between the nodule and the pleura and ribs, which serves as the angular information of the pulmonary nodule.
[0063] The angle and depth information of the virtual nodule is converted into physical coordinates of the actual lung surface. This allows the virtual lung nodule to be precisely superimposed onto the corresponding location on the patient's body surface when the user observes it through an AR device (such as a head-mounted display or surgical navigation probe), while simultaneously displaying both depth and angle information. Preferably, depth information can be displayed using color gradients, and angle information can be displayed using arrows indicating the angular direction.
[0064] This invention uses a 3D model of the diseased lung before surgery, CT data, and acquired images of the diseased lung to generate a 3D model of the lung lobe where the lesion is located during surgery using AI and 3D reconstruction technology. Augmented reality (AR) technology is then used to overlay the image data onto the corresponding lung area in real time during surgery. Precise matching is achieved between the lung and the virtual lung to pinpoint the depth and angle of the lung nodule, providing an intuitive localization experience.
[0065] S6 tracks and locates pulmonary nodules in real time based on the activity of the patient's solid lungs; Real-time tracking and labeling guidance of the pulmonary nodules; In one embodiment of this specification, a motion tracking algorithm is used to track the displacement of lung tissue in intraoperative thoracoscopic images (such as changes in nodule position caused by respiration) in real time, and to synchronously update the coordinates of the intraoperative dynamic 3D model of the lung lobe where the lesion is located.
[0066] Based on the updated intraoperative dynamic 3D model, the AR screen prompts the user to mark the location on the patient's body surface (such as projecting a virtual point), guiding the user to use a special marking pen to mark the patient's solid lung surface. This invention locates pulmonary nodules intraoperatively and tracks the location points in real time based on lung activity, providing markers for doctors to mark on the solid lung, facilitating subsequent resection of the pulmonary nodules.
[0067] With this invention, the location of pulmonary nodules does not require the placement of a positioning pin before surgery, reducing patient pain and waiting time. No re-opening is required during surgery, and the entire positioning process will not cause harm to the patient.
[0068] Figure 2 This is a schematic diagram of a lung nodule-assisted localization system provided in an embodiment of this specification. The system includes: Preoperative diseased lung modeling module 210 is used to construct a virtual model of the patient's preoperative lungs; Thoracoscopic video input module 220 is used to acquire images of the lesion through a thoracoscope; The intraoperative lung modeling module 230 is used to generate an intraoperative dynamic 3D model based on the preoperative lung virtual model and the lesion image. Video acquisition module 240 is used to acquire real-time intraoperative images through the thoracoscope; The intelligent positioning module 250 is used to combine the intraoperative dynamic 3D model with the intraoperative real-time images to help predict the location of pulmonary nodules in the patient's solid lung.
[0069] Optionally, the preoperative diseased lung modeling module 210 includes: The data acquisition submodule is used to acquire the patient's preoperative CT data; The three-dimensional reconstruction submodule is used to perform three-dimensional reconstruction based on the preoperative CT data and generate the preoperative lung virtual model. The tagging submodule is used to tag virtual lung nodules in the preoperative lung virtual model.
[0070] Optionally, the intraoperative lung modeling module 230 includes: The alignment submodule is used to spatially align the preoperative lung virtual model with the lesion image; The model building submodule is used to generate an intraoperative dynamic 3D model of the lung lobe where the lesion is located based on the spatial alignment results and the patient's respiratory movements.
[0071] Optionally, the alignment submodule includes: The marker lookup unit is used to find common markers between the lesion image and the preoperative lung virtual model; Alignment unit, used to align the lesion image with the preoperative lung virtual model based on the common marker.
[0072] Optionally, the intelligent positioning module 250 includes: The image acquisition submodule is used to acquire real-time images of the patient's actual lung region based on the intraoperative real-time images; The overlay and fusion submodule is used to fuse the intraoperative dynamic 3D model with the actual lung region using the real-time image; The positioning submodule is used to assist in locating the position of the pulmonary nodule in the patient's physical lung based on the depth and angle information of the virtual pulmonary nodule relative to the patient's body surface.
[0073] Optional, also includes: The lung nodule real-time location tracking module is used to track and locate lung nodules in real time based on the activity of the patient's solid lungs.
[0074] The functions of the system in this embodiment have been described in the above method embodiments. Therefore, for any parts not detailed in this embodiment, please refer to the relevant descriptions in the foregoing embodiments, which will not be repeated here.
[0075] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification. The electronic device includes a memory 301 and a processor 302. The memory 301 is used to store computer-executable instructions. When the computer-executable instructions are executed by the processor 302, they can implement the steps of the above-described method embodiments.
[0076] Figure 4 This is a schematic diagram of the structure of a computer-readable storage medium 400 provided in an embodiment of this specification. The computer-readable storage medium 400 stores one or more computer programs, which, when executed by a processor, can implement the steps of the above-described method embodiments.
[0077] This specification also provides a computer program product, including a computer program / computer executable instructions, which, when executed by a processor, can implement the steps of the above-described method embodiments.
[0078] Those skilled in the art will understand that all or part of the processes in the above method embodiments can be implemented by a computer program instructing related hardware. When the computer program is executed, it may include the processes of the embodiments of the above methods.
[0079] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for assisted localization of pulmonary nodules, characterized in that, include: Construct a preoperative virtual model of the patient's lungs; Images of the lesion were acquired via thoracoscopy; Based on the preoperative lung virtual model and the lesion image, an intraoperative dynamic 3D model is generated; Real-time intraoperative images are acquired using the thoracoscopic technique; By combining the intraoperative dynamic 3D model with the intraoperative real-time images, the location of pulmonary nodules in the patient's solid lung can be predicted.
2. The method for assisted localization of pulmonary nodules as described in claim 1, characterized in that, The construction of the patient's preoperative lung virtual model includes: Obtain the patient's preoperative CT data; Based on the preoperative CT data, a three-dimensional reconstruction is performed to generate the preoperative lung virtual model; Virtual lung nodules are marked in the preoperative virtual lung model.
3. The method for assisted localization of pulmonary nodules as described in claim 2, characterized in that, The step of generating an intraoperative dynamic 3D model of the lung lobe where the lesion is located based on the preoperative virtual lung model and the lesion image includes: Spatially align the preoperative lung virtual model with the lesion image; Based on the spatial alignment results and the patient's respiratory movements, an intraoperative dynamic 3D model of the lung lobe where the lesion is located is generated.
4. The method for assisted localization of pulmonary nodules as described in claim 3, characterized in that, The step of spatially aligning the preoperative lung virtual model with the lesion image includes: Identify common features between the lesion image and the preoperative lung virtual model; The lesion image is aligned with the preoperative lung virtual model based on the common markers.
5. The method for assisted localization of pulmonary nodules as described in claim 3, characterized in that, The method of combining the intraoperative dynamic 3D model with the intraoperative real-time images to assist in predicting the location of pulmonary nodules in the patient's solid lung includes: Real-time images of the patient's actual lung region are obtained based on the intraoperative real-time images; The intraoperative dynamic 3D model is fused with the actual lung area using the real-time video feed. Based on the depth and angle information of the virtual lung nodule relative to the patient's body surface, the location of the lung nodule in the patient's physical lung is assisted.
6. The method for assisted localization of pulmonary nodules as described in claim 5, characterized in that, Also includes: Lung nodules are located in real time based on the activity of the patient's solid lungs.
7. A lung nodule-assisted localization system, characterized in that, include: The preoperative diseased lung modeling module is used to construct a virtual model of the patient's preoperative lungs; Thoracoscopic video input module, used to acquire images of the lesion through thoracoscopy; The intraoperative lung modeling module is used to generate an intraoperative dynamic 3D model based on the preoperative virtual lung model and the lesion image. The video acquisition module is used to acquire real-time intraoperative images through the thoracoscope; The intelligent positioning module is used to combine the intraoperative dynamic 3D model with the intraoperative real-time images to help predict the location of pulmonary nodules in the patient's solid lung.
8. A computer device, characterized in that, The computer device includes: Processor; and, A memory storing computer-executable instructions, which, when executed, cause the processor to perform the method of any one of claims 1-6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs / instructions, which, when executed by a processor, implement the method as described in any one of claims 1-6.
10. A computer program product, characterized in that, Includes a computer program / instruction, which, when executed by a processor, implements the method as described in any one of claims 1-6.