A complex chest trauma surgery auxiliary stereoscopic imaging system and method

Abstract

The present application belongs to the technical field of medical image auxiliary, and specifically relates to a complex chest trauma surgery auxiliary stereoscopic imaging system and method. The present application realizes accurate assistance to the thymus resection surgery process through integration of multi-modal image data, real-time three-dimensional space capture, dynamic correction and view adjustment driven by risk score, not only improves the accuracy and safety of the surgery, but also significantly reduces the surgery risk and operation difficulty. Through real-time acquisition and processing of multi-modal image data, a high-precision three-dimensional anatomical structure model containing thymus tissue, blood vessel network and nerve distribution can be constructed, providing an intuitive surgery navigation view for doctors. Meanwhile, a dynamic correction technology is adopted, the thymus displacement parameters caused by the respiratory movement of the patient's chest cavity are monitored in real time, and the real-time three-dimensional space data is dynamically adjusted based on the parameters, ensuring the spatial synchronization of the image data and the anatomical structure, and effectively avoiding the navigation error caused by the respiratory movement.

Description

Technical Field

[0001] This invention belongs to the field of medical imaging-assisted technology, specifically relating to a three-dimensional imaging system and method for assisting in complex chest trauma surgery. Background Technology

[0002] With the continuous advancement of medical technology, the requirements for precision in surgical procedures are increasing, especially in complex thymectomy. Traditional surgical navigation methods often rely on two-dimensional imaging data, which makes it difficult to intuitively and in real time reflect the three-dimensional anatomical structure of the surgical area, thereby increasing surgical risks and operational difficulties. With the rapid development of three-dimensional imaging technology, its application in the medical field is becoming more and more widespread. Surgical navigation based on three-dimensional imaging technology can significantly improve the accuracy and safety of surgery.

[0003] However, existing three-dimensional imaging technology still has some problems in thymectomy, such as real-time updating of image data, accurate reconstruction of anatomical structures, and dynamic changes in anatomical structures during surgery. These problems often lead to insufficient real-time performance and accuracy of surgical navigation, which in turn affects the surgical outcome. In addition, there are areas that may be obscured by instruments during the operation, which may limit the surgeon's field of vision and affect the precision and safety of the operation. To solve the above problems, this invention proposes a three-dimensional imaging method to assist in complex chest trauma surgery, thereby improving the precision and safety of the operation. Summary of the Invention

[0004] The purpose of this invention is to provide a three-dimensional imaging system and method for assisting in complex chest trauma surgery, which can acquire multimodal image data of the thymus region in real time, construct an accurate three-dimensional anatomical structure model, and dynamically correct the spatial synchronization between the image data and the anatomical structure to generate a three-dimensional navigation view to assist doctors in performing surgical operations.

[0005] The specific technical solution adopted by this invention is as follows:

[0006] A method for assisting in the stereoscopic imaging of complex chest trauma surgery includes:

[0007] Acquire multimodal imaging data of the thymus region, including CT images and ultrasound images;

[0008] A three-dimensional anatomical model containing thymus tissue, vascular network, and nerve distribution was constructed using image segmentation processing.

[0009] Real-time capture of visible light images and near-infrared spectral data of the surgical field to generate real-time three-dimensional spatial data;

[0010] Real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements were collected, and the real-time three-dimensional spatial data were dynamically corrected based on the real-time displacement parameters to maintain spatial synchronization between the imaging data and the anatomical structure.

[0011] The corrected three-dimensional anatomical model is overlaid on the surgical field in real time to generate a stereoscopic navigation view, and the display angle, depth and transparency of the stereoscopic navigation view are dynamically adjusted according to the risk score of the resection location.

[0012] In a preferred embodiment, after the multimodal image data acquisition is completed, preprocessing is performed simultaneously. The preprocessing steps include:

[0013] Isotropic sampling processing is performed on CT images to uniformly adjust the slice thickness to a preset standard value;

[0014] Dynamic range compression and speckle noise suppression are performed on ultrasound images, and speckle noise is eliminated by anisotropic diffusion filtering;

[0015] CT and ultrasound images were acquired at the end of expiration and peak of inspiration, respectively, and the CT and ultrasound images were fused to correct for respiratory artifacts.

[0016] In a preferred embodiment, the step of constructing a three-dimensional anatomical model including thymus tissue, vascular network, and neural distribution through image segmentation processing includes:

[0017] CT images were processed by image feature comparison and combined with the texture features of ultrasound images for auxiliary segmentation to extract the contour of thymus tissue.

[0018] By identifying and marking the main blood vessels of a preset diameter, and combining this with ultrasound blood flow imaging to confirm the direction of blood vessel branches, the capillary network can be supplemented.

[0019] Based on the high-resolution characteristics of ultrasound images, the direction of nerve fibers is identified and a nerve distribution map is generated.

[0020] The segmentation results from various image sources are spatially matched, and the multimodal image data are fused through respiratory cycle synchronization correction to generate a three-dimensional anatomical model of the thymus region.

[0021] In a preferred embodiment, the step of real-time capture of visible light images and near-infrared spectral data of the surgical field to generate real-time three-dimensional spatial data includes:

[0022] Dual-channel synchronous imaging is used to acquire high-resolution visible light images and near-infrared spectral information of the surgical field. The dual channels include a visible light channel and a near-infrared channel. The visible light channel is equipped with a polarizing filter, and the near-infrared channel is set with a specific wavelength range to enhance the contrast of vascular imaging.

[0023] Dynamic threshold segmentation of near-infrared spectral information is performed to extract blood vessel contours, and spatial registration is performed in conjunction with visible light images to align blood vessels with tissue structures.

[0024] The registered visible light image is fused with near-infrared spectral information and input into a pre-trained deep learning model to predict the anatomical topological relationships of the instrument-occluded area within the surgical field, generating real-time three-dimensional spatial data with confidence weights.

[0025] In a preferred embodiment, the step of predicting the anatomical topological relationships of the instrument-obstructed area within the surgical field and generating real-time three-dimensional spatial data with confidence weights includes:

[0026] The instrument's position is acquired from multiple consecutive frames of near-infrared spectral information, and spatial mapping is performed based on the instrument's position to establish the instrument's movement trajectory;

[0027] Based on the instrument's historical movement trajectory and current breathing stage, predict the potential area of ​​anatomical structure obstruction by the instrument at the next moment;

[0028] Anatomical structures within potentially occluded areas are hierarchically labeled, and corresponding confidence weights are determined based on the degree of occlusion. These confidence weights are then encoded as transparency parameters, which are then integrated into the real-time 3D spatial data model.

[0029] In a preferred embodiment, the step of acquiring real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements includes:

[0030] By pre-deploying a miniature sensor array in the patient's thoracic cavity, the amplitude and frequency of intercostal muscle contraction can be monitored in real time to obtain thymus displacement data;

[0031] By selecting respiratory motion feature markers in CT images and combining them with real-time diaphragmatic displacement data from ultrasound images, a three-dimensional spatial displacement vector field is constructed.

[0032] Collect real-time and historical displacement data of the patient's current respiratory cycle and historical respiratory cycles, perform time series analysis, and output predicted deformation parameters of thymus tissue;

[0033] The real-time displacement parameters of the thymus are generated by superimposing the three-dimensional spatial displacement vector field with the predicted deformation parameters.

[0034] In a preferred embodiment, the step of dynamically correcting the real-time three-dimensional spatial data based on real-time displacement parameters to maintain spatial synchronization between the image data and the anatomical structure includes:

[0035] Based on the three-dimensional spatial displacement vector field and predicted deformation parameters, the coordinate compensation amount of real-time three-dimensional spatial data is calculated, and the image data is adjusted frame by frame according to the coordinate compensation amount.

[0036] The three-dimensional coordinates of the thoracic anatomical structure are updated at a preset update interval during the respiratory cycle by synchronous acquisition of visible light and near-infrared channels.

[0037] Based on the real-time monitoring values ​​fed back by the micro-sensor array, the predicted deformation parameters are compared in real time to output the spatial deviation, and the coordinate compensation amount is dynamically corrected according to the spatial deviation.

[0038] When the spatial deviation exceeds the preset threshold, an alarm mechanism is triggered, and an ultrasound scan is initiated for secondary calibration.

[0039] In a preferred embodiment, the step of dynamically adjusting the display angle, depth, and transparency of the stereoscopic navigation view based on the risk score of the excision location includes:

[0040] Calculate the Euclidean distance between the current resection point and the vascular network in the three-dimensional anatomical model, and record it as the first conditional parameter;

[0041] Collect the minimum distance between the current resection point and the nerve fiber bundle, and record it as the second conditional parameter;

[0042] The first and second conditional parameters are weighted and fused to generate a risk score.

[0043] The display angle, depth, and transparency of the 3D navigation view are dynamically adjusted based on the risk score.

[0044] When the risk score exceeds the preset risk threshold, an ultrasound scan is triggered for secondary calibration, and the topology of the obscured area is updated simultaneously.

[0045] When the risk score is lower than the preset risk threshold, the risk deviation value is output, and the display parameters of the 3D navigation view are fine-tuned based on the risk deviation value.

[0046] The present invention also provides a surgical-assisted stereoscopic imaging system for thymectomy, using the above-described method for assisting stereoscopic imaging in complex chest trauma surgery, comprising:

[0047] The data acquisition module is used to acquire multimodal imaging data of the thymus region, including CT images and ultrasound images.

[0048] The model building module is used to construct a three-dimensional anatomical model containing thymus tissue, vascular network, and neural distribution through image segmentation processing.

[0049] The surgical field capture module is used to capture visible light images and near-infrared spectral data of the surgical field in real time and generate real-time three-dimensional spatial data.

[0050] The correction module is used to collect real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements, and to dynamically correct the real-time three-dimensional spatial data based on the real-time displacement parameters to maintain spatial synchronization between the imaging data and the anatomical structure.

[0051] The view adjustment module is used to overlay the corrected three-dimensional anatomical structure model with the surgical field in real time to generate a stereo navigation view, and dynamically adjust the display angle, depth and transparency of the stereo navigation view according to the risk score of the resection location.

[0052] And, an electronic device, the electronic device comprising:

[0053] At least one processor;

[0054] and a memory communicatively connected to the at least one processor;

[0055] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the aforementioned method for assisting in the complex chest trauma surgery with stereoscopic imaging.

[0056] The technical effects achieved by this invention are as follows:

[0057] This invention achieves precise assistance in the thymectomy procedure by integrating multimodal image data, real-time 3D spatial capture, dynamic correction, and risk-scoring-driven view adjustment. This method not only improves the accuracy and safety of the surgery but also significantly reduces surgical risks and operational difficulty. By acquiring and processing multimodal image data in real time, this invention can construct a high-precision 3D anatomical model including thymic tissue, vascular networks, and nerve distribution, providing surgeons with an intuitive surgical navigation view. This allows surgeons to clearly identify anatomical structures during surgery, enabling more accurate surgical decisions. Furthermore, dynamic correction technology is employed, which monitors thymic displacement parameters caused by the patient's thoracic respiratory movements in real time and dynamically adjusts the real-time 3D spatial data accordingly. This ensures spatial synchronization between the image data and the anatomical structure, effectively avoiding navigation errors caused by respiratory movements. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0059] Figure 2 This is a schematic diagram of the system modules of the present invention;

[0060] Figure 3 This is a schematic diagram of the electronic device structure of the present invention. Detailed Implementation

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in a preferred embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that mutually excludes other embodiments.

[0064] Please see Figure 1 As shown, the present invention provides a method for assisting in three-dimensional imaging during complex chest trauma surgery, comprising:

[0065] S1. Acquire multimodal imaging data of the thymus region, including CT images and ultrasound images;

[0066] In step S1, before the thymectomy, multimodal imaging data of the thymus region is acquired, including computed tomography (CT) images and ultrasound images. CT images provide high-resolution anatomical information, while ultrasound images can display the real-time dynamics of vascular networks and nerve distribution. The combination of the two provides rich information for three-dimensional reconstruction. After the multimodal imaging data acquisition is completed, preprocessing is performed simultaneously. The preprocessing steps include:

[0067] Isotropic sampling processing is performed on CT images to uniformly adjust the slice thickness to a preset standard value;

[0068] Dynamic range compression and speckle noise suppression are performed on ultrasound images, and speckle noise is eliminated by anisotropic diffusion filtering;

[0069] CT images and ultrasound images were acquired at the end of expiration and the peak of inspiration, respectively, and the CT images and ultrasound images were fused to correct for respiratory artifacts.

[0070] Specifically, after the multimodal image data acquisition is completed, in order to improve the quality of the image data and provide a more accurate foundation for subsequent 3D reconstruction and surgical navigation, the CT images are first subjected to isotropic sampling processing. During this stage, the slice thickness of the CT images is uniformly adjusted to a preset standard value, such as 1 mm or 2 mm, to ensure uniform resolution across the entire volumetric data. This processing method facilitates seamless image browsing and analysis in 3D space, especially for cases requiring precise measurements or 3D reconstruction. For ultrasound images, the preprocessing steps include dynamic range compression and speckle noise suppression. The dynamic range of ultrasound images is typically large, which may lead to bright areas and... Details in dark areas cannot be clearly displayed simultaneously. Dynamic range compression can adjust the image contrast to make the boundaries of different tissues more distinct. At the same time, since speckle noise is a common problem in ultrasound imaging, the application of anisotropic diffusion filters can effectively eliminate irregular speckles and improve image clarity. Subsequently, CT images and ultrasound images are acquired at two key breathing points—end-expiration and peak-inspiratory phases, respectively. This is mainly because respiratory movements may cause slight changes in the position of internal organs, producing respiratory artifacts. By comparing images from different breathing phases and performing fusion processing, positional deviations caused by breathing can be corrected, thereby providing more accurate comparison and positioning information.

[0071] S2. Construct a three-dimensional anatomical model containing thymus tissue, vascular network, and nerve distribution through image segmentation processing;

[0072] In step S2, after the multimodal image data processing is completed, the processed multimodal image data is finely segmented using appropriate image segmentation techniques to construct a three-dimensional anatomical model containing thymus tissue, vascular network, and nerve distribution. This provides a precise anatomical reference for thymectomy surgical planning. The step of constructing the three-dimensional anatomical model containing thymus tissue, vascular network, and nerve distribution through image segmentation includes:

[0073] CT images were processed by image feature comparison and combined with the texture features of ultrasound images for auxiliary segmentation to extract the contour of thymus tissue.

[0074] By identifying and marking the main blood vessels of a preset diameter, and combining this with ultrasound blood flow imaging to confirm the direction of blood vessel branches, the capillary network can be supplemented.

[0075] Based on the high-resolution characteristics of ultrasound images, the direction of nerve fibers is identified and a nerve distribution map is generated.

[0076] The segmentation results of each image source are spatially matched and corrected synchronously through the respiratory cycle to achieve the fusion of multimodal image data and generate a three-dimensional anatomical model of the thymus region.

[0077] Specifically, in constructing the three-dimensional anatomical model, image feature comparison technology is first used to process CT images to enhance the boundaries of the thymus tissue, making it stand out more against a high-contrast background. Simultaneously, by introducing texture features from ultrasound images, more precise segmentation is assisted, extracting the accurate contour of the thymus tissue and ensuring the initial accuracy of the model. Then, vascular recognition technology is used to automatically label the main blood vessels of a preset diameter, making the main vessels clearly visible in the CT images. Subsequently, combined with dynamic information from ultrasound blood flow imaging, the direction of vascular branches can be tracked and confirmed, thus constructing a complete vascular network. After identifying the vascular structure, the direction of nerve fibers can be identified by analyzing ultrasound images, forming a nerve distribution atlas. Finally, the segmentation results from CT, ultrasound, and nerve distribution are spatially matched. This process requires synchronous correction of the respiratory cycle for different image sources to eliminate anatomical displacement caused by respiratory movements, ensuring that each structure can be accurately aligned in three-dimensional space, thereby generating a complete three-dimensional anatomical model of the thymus region.

[0078] S3. Real-time capture of visible light images and near-infrared spectral data of the surgical field to generate real-time three-dimensional spatial data;

[0079] In step S3, during the surgery, visible light images and near-infrared spectral data of the surgical field are captured in real time to generate real-time three-dimensional spatial data. Near-infrared spectroscopy allows for tracking the real-time displacement of the thymus caused by thoracic respiratory movements, thereby maintaining synchronization between the image and the actual anatomical structure. The step of capturing visible light images and near-infrared spectral data of the surgical field in real time to generate real-time three-dimensional spatial data includes:

[0080] Dual-channel synchronous imaging is used to acquire high-resolution visible light images and near-infrared spectral information of the surgical field. The dual channels include a visible light channel and a near-infrared channel. The visible light channel is equipped with a polarizing filter, and the near-infrared channel is set with a specific wavelength range to enhance the contrast of vascular imaging.

[0081] Dynamic threshold segmentation of near-infrared spectral information is performed to extract blood vessel contours, and spatial registration is performed in conjunction with visible light images to align blood vessels with tissue structures.

[0082] The registered visible light image is fused with near-infrared spectral information and input into a pre-trained deep learning model to predict the anatomical topological relationship of the instrument-occluded area in the surgical field and generate real-time three-dimensional spatial data with confidence weights.

[0083] Specifically, during the surgery, high-resolution visible light images and near-infrared spectral information of the surgical field are acquired through dual-channel imaging. The visible light channel is equipped with a polarizing filter to reduce interference from ambient light and enhance image contrast, making tissue structures clearer. The near-infrared channel is set within a specific wavelength range. The purpose of setting the wavelength range is to maximize the absorption of near-infrared light by hemoglobin in blood vessels, thereby improving the effect of vascular visualization. Then, the near-infrared spectral information is processed by dynamic threshold segmentation to identify the vascular contour, and the segmented vascular contour is spatially registered with the visible light image (e.g., through algorithms such as feature point matching) to ensure the accuracy of vascular visualization. Aligning the surgical field with other tissue structures provides an accurate reference for subsequent 3D reconstruction. After registration, data that integrates visible light image information and near-infrared spectral information is input into a pre-trained deep learning model (specifically, trained and optimized based on existing image data using known deep learning frameworks such as convolutional neural networks). This model can predict the topological relationships of anatomical structures in the instrument-occluded area. Through the prediction of the deep learning model, the generated 3D spatial data not only contains the structural information of the surgical field but also includes confidence weights. The confidence weights can indicate the accuracy of the deep learning model's predictions, providing real-time decision support for doctors.

[0084] Secondly, the steps of predicting the anatomical topological relationships of the instrument-obstructed area within the surgical field and generating real-time three-dimensional spatial data with confidence weights include:

[0085] The instrument's position is acquired from multiple consecutive frames of near-infrared spectral information, and spatial mapping is performed based on the instrument's position to establish the instrument's movement trajectory;

[0086] Based on the instrument's historical movement trajectory and current breathing stage, predict the potential area of ​​anatomical structure obstruction by the instrument at the next moment;

[0087] Anatomical structures within potentially occluded areas are hierarchically labeled, and corresponding confidence weights are determined based on the degree of occlusion. The confidence weights are then encoded as transparency parameters, which are then integrated into the real-time 3D spatial data model.

[0088] In this implementation, multiple frames of near-infrared spectral information are first continuously acquired. This near-infrared spectral information accurately captures the instrument's position in the surgical field and performs real-time spatial mapping to construct the instrument's dynamic movement trajectory. This is similar to drawing the instrument's movement path in virtual space, providing a foundation for subsequent predictive analysis. Then, the instrument's historical movement trajectory is used to predict its future movement trend. Here, considering the influence of physiological activities such as respiration, the movement pattern of the instrument during the current respiratory phase is analyzed to predict the area of ​​anatomical obstruction that the instrument may cause to the anatomical structures at the next moment. This helps doctors anticipate potential visual obstructions and make appropriate decisions. In accordance with the corresponding operational adjustments, after identifying potential occlusion areas, the anatomical structures within these areas are graded and labeled. Based on the severity of the occlusion, such as complete occlusion, partial occlusion, or only causing light interference, different confidence weights are assigned to the anatomical structures. These confidence weights help doctors quickly identify the severity of the occlusion and prioritize the treatment of important or urgent areas. In this embodiment, the confidence weights are converted into transparency parameters; areas with higher occlusion levels have higher transparency. Thus, in the three-dimensional spatial data model, doctors can observe which areas may disappear behind instruments and which areas remain visible, thereby enhancing the practicality and intuitiveness of the real-time three-dimensional view.

[0089] S4. Collect real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements, and dynamically correct the real-time three-dimensional spatial data based on the real-time displacement parameters to maintain spatial synchronization between the imaging data and the anatomical structure.

[0090] In step S4, the three-dimensional spatial data is dynamically corrected based on real-time displacement parameters to eliminate image drift caused by the patient's respiratory movements. This ensures that the image data remains consistent with the anatomical structures, providing the surgeon with stable and reliable surgical field navigation. The step of acquiring real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements includes:

[0091] By pre-deploying a miniature sensor array in the patient's thoracic cavity, the amplitude and frequency of intercostal muscle contraction can be monitored in real time to obtain thymus displacement data;

[0092] By selecting respiratory motion feature markers in CT images and combining them with real-time diaphragmatic displacement data from ultrasound images, a three-dimensional spatial displacement vector field is constructed.

[0093] Collect real-time and historical displacement data of the patient's current respiratory cycle and historical respiratory cycles, perform time series analysis, and output predicted deformation parameters of thymus tissue;

[0094] The three-dimensional spatial displacement vector field is superimposed with the predicted deformation parameters to generate the real-time displacement parameters of the thymus.

[0095] Specifically, when collecting real-time displacement parameters of the patient's thymus, a miniature sensor array is first deployed in the patient's thoracic cavity to monitor the contraction amplitude and frequency of the intercostal muscle group in real time, thus effectively tracking the minute displacements of the thymus. To construct a more accurate three-dimensional spatial model, characteristic markers of respiratory motion are selected in CT images. These markers, combined with real-time diaphragmatic displacement data provided by ultrasound images, form a dynamic three-dimensional spatial displacement vector field, marking the trajectory of the thymus during each breath. Then, displacement data from the patient's current respiratory cycle and several past respiratory cycles are collected and subjected to time series analysis (such as using time series prediction algorithms like autoregressive integral moving average models or long short-term memory networks) to identify the deformation patterns of the thymus tissue and output its predicted deformation parameters at different respiratory stages. Finally, the obtained three-dimensional spatial displacement vector field is superimposed with the predicted deformation parameters (specifically, each coordinate point of the three-dimensional spatial displacement vector field is multiplied by the predicted deformation parameter and weighted summed). The resulting real-time displacement parameters comprehensively demonstrate the dynamic changes of the thymus during respiration.

[0096] Secondly, the steps of dynamically correcting real-time three-dimensional spatial data based on real-time displacement parameters to maintain spatial synchronization between image data and anatomical structures include:

[0097] Based on the three-dimensional spatial displacement vector field and predicted deformation parameters, the coordinate compensation amount of real-time three-dimensional spatial data is calculated, and the image data is adjusted frame by frame according to the coordinate compensation amount.

[0098] The three-dimensional coordinates of the thoracic anatomical structure are updated at a preset update interval during the respiratory cycle by synchronous acquisition of visible light and near-infrared channels.

[0099] Based on the real-time monitoring values ​​fed back by the micro-sensor array, the predicted deformation parameters are compared in real time to output the spatial deviation, and the coordinate compensation amount is dynamically corrected according to the spatial deviation.

[0100] When the spatial deviation exceeds the preset threshold, an alarm mechanism is triggered, and an ultrasound scan is initiated for secondary calibration.

[0101] Specifically, when adjusting real-time 3D spatial data using real-time displacement parameters, the coordinate compensation amount for each frame of image data is first calculated based on the 3D spatial displacement vector field and the predicted deformation parameters. The formula for calculating the coordinate compensation amount is as follows: In the formula, Indicates the coordinate compensation amount. Represents a three-dimensional spatial coordinate point. Indicates time parameters (respiratory cycle phases). Indicates the first Three-dimensional spatial displacement vector field components The weighting coefficients represent the displacement vector field. This represents the total number of field components of the displacement vector, with a value of 3, corresponding to... direction, This represents the correction coefficient for the predicted deformation parameters. This system represents predicted deformation parameters to compensate for minute displacements caused by physiological movements. In this way, multimodal imaging data can be adjusted to correspond to the actual anatomical structure. Then, it is simultaneously acquired through visible light and near-infrared channels, and the three-dimensional coordinates of the thoracic anatomical structure are updated at preset update intervals within the respiratory cycle. This provides more comprehensive and real-time information on changes in anatomical structures. The data monitored in real time by the micro-sensor array is then compared with the predicted deformation parameters. When spatial deviation is detected, it is immediately output, and the coordinate compensation amount is dynamically corrected accordingly. This ensures that the correction process can be maintained even under rapidly changing physiological conditions. It should be noted that when the spatial deviation exceeds the preset deviation threshold, an alarm mechanism is triggered to alert the operator. At the same time, an ultrasound scan is initiated for secondary calibration to avoid navigation inaccuracies caused by error accumulation.

[0102] S5. The corrected three-dimensional anatomical structure model is superimposed on the surgical field in real time to generate a stereo navigation view, and the display angle, depth and transparency of the stereo navigation view are dynamically adjusted according to the risk score of the resection location.

[0103] In step S5, the corrected three-dimensional anatomical model is overlaid with the real-time surgical field image to generate a stereoscopic navigation view. Based on the risk score of the resection location, the display angle, depth, and transparency are dynamically adjusted, allowing surgeons to more intuitively identify high-risk areas and make accurate surgical decisions. The step of dynamically adjusting the display angle, depth, and transparency of the stereoscopic navigation view based on the risk score of the resection location includes:

[0104] Calculate the Euclidean distance between the current resection point and the vascular network in the three-dimensional anatomical model, and record it as the first conditional parameter;

[0105] Collect the minimum distance between the current resection point and the nerve fiber bundle, and record it as the second conditional parameter;

[0106] The first and second conditional parameters are weighted and fused to generate a risk score.

[0107] The display angle, depth, and transparency of the 3D navigation view are dynamically adjusted based on the risk score.

[0108] When the risk score exceeds the preset risk threshold, an ultrasound scan is triggered for secondary calibration, and the topology of the obscured area is updated simultaneously.

[0109] When the risk score is lower than the preset risk threshold, the risk deviation value is output, and the display parameters of the 3D navigation view are fine-tuned based on the risk deviation value.

[0110] Specifically, the process first calculates the Euclidean distance between the current resection point and the vascular network in the 3D anatomical model. This distance is the straight-line distance from the resection point to the nearest blood vessel in 3D space, quantifying the potential impact of the resection on the blood vessel. This distance is recorded as the first conditional parameter. To protect nerve fiber bundles, the minimum distance between the current resection point and sensitive structures is also collected and recorded as the second conditional parameter. The first and second conditional parameters are then weighted and fused to generate a corresponding risk score. Based on the risk score, the display parameters of the stereo navigation view are adjusted in real time. If the risk score exceeds a preset risk threshold, it indicates an increased potential threat to surrounding tissues. In this case, an ultrasound scan is triggered for secondary calibration, which helps to further confirm the safety of the resection point and simultaneously updates the topology of the occluded area to ensure the accuracy of the navigation information. When the risk score is below the preset risk threshold, it indicates a relatively low surgical risk. In this case, a risk deviation value (the difference between the risk threshold and the risk score) is output, and the display parameters of the stereo navigation view are fine-tuned based on the risk deviation value to ensure that the surgeon receives appropriate visual guidance.

[0111] Please see Figure 2 A thymectomy surgical aid stereoscopic imaging system, using the aforementioned complex chest trauma surgical aid stereoscopic imaging method, includes:

[0112] The data acquisition module is used to acquire multimodal imaging data of the thymus region, including CT images and ultrasound images.

[0113] The model building module is used to construct a three-dimensional anatomical model containing thymus tissue, vascular network, and neural distribution through image segmentation processing.

[0114] The surgical field capture module is used to capture visible light images and near-infrared spectral data of the surgical field in real time and generate real-time three-dimensional spatial data.

[0115] The correction module is used to collect real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements, and to dynamically correct the real-time three-dimensional spatial data based on the real-time displacement parameters to maintain spatial synchronization between the imaging data and the anatomical structure.

[0116] The view adjustment module is used to overlay the corrected three-dimensional anatomical structure model with the surgical field in real time to generate a stereo navigation view, and dynamically adjust the display angle, depth and transparency of the stereo navigation view according to the risk score of the resection location.

[0117] As described above, the data acquisition module can accurately acquire multimodal imaging data of the thymus region, including high-precision CT images and real-time ultrasound images, laying the foundation for subsequent model construction. The model construction module uses advanced image segmentation processing technology to construct a three-dimensional anatomical structure model containing thymus tissue, vascular network, and nerve distribution. The surgical field capture module is responsible for capturing visible light images and near-infrared spectral data of the surgical field in real time. After processing, the visible light images and near-infrared spectral data can generate real-time three-dimensional spatial data, providing real-time information for dynamic correction during the operation. The correction module can collect real-time displacement parameters of the thymus caused by the patient's chest cavity respiratory movements and dynamically correct the real-time three-dimensional spatial data based on the real-time displacement parameters to ensure that the imaging data and anatomical structure are spatially synchronized, thereby improving the accuracy and safety of the surgical process. The view adjustment module is responsible for superimposing the corrected three-dimensional anatomical structure model with the surgical field in real time to generate a stereoscopic navigation view. It also dynamically adjusts the display angle, depth, and transparency of the stereoscopic navigation view according to the risk score of the resection location, enabling doctors to more intuitively identify high-risk areas and make more accurate surgical decisions.

[0118] Please see Figure 3 An electronic device, comprising:

[0119] At least one processor;

[0120] and memory that is communicatively connected to at least one processor;

[0121] The memory stores a computer program that can be executed by at least one processor, which enables the at least one processor to perform the aforementioned method for assisting in the complex chest trauma surgery with stereoscopic imaging.

[0122] The processor of the aforementioned electronic device can be a central processing unit (CPU), a graphics processing unit (GPU), or a digital signal processor (DSP). The memory can include read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or solid-state drive (SSD). The computer program stored in the memory can be executed by the processor to implement the aforementioned thymectomy surgical-assisted stereoscopic imaging method. In addition, the electronic device may also include components such as an arithmetic logic unit (ALU), input devices, output devices, and network interfaces. The ALU can provide computational support for at least one of the CPU, GPU, or DSP. Input devices, such as a keyboard, mouse, or touch screen, are used to receive user input commands. Output devices, such as a monitor or printer, are used to display processing results or print output. The network interface is responsible for communication connections between the electronic device and other devices, ensuring data transmission and sharing.

[0123] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0124] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A three-dimensional imaging system for assisting in complex chest trauma surgery, characterized in that... include: The data acquisition module is used to acquire multimodal imaging data of the thymus region, including CT images and ultrasound images. The model building module is used to construct a three-dimensional anatomical model containing thymus tissue, vascular network, and neural distribution through image segmentation processing. The surgical field capture module is used to capture visible light images and near-infrared spectral data of the surgical field in real time and generate real-time three-dimensional spatial data. The registered visible light images and near-infrared spectral information are fused and input into a pre-trained deep learning model to predict the anatomical topological relationship of the instrument-occluded area in the surgical field and generate real-time three-dimensional spatial data with confidence weights. The step of generating real-time three-dimensional spatial data with confidence weights includes: The instrument's position is acquired from multiple consecutive frames of near-infrared spectral information, and spatial mapping is performed based on the instrument's position to establish the instrument's movement trajectory; Based on the instrument's historical movement trajectory and current breathing stage, predict the potential area of ​​anatomical structure obstruction by the instrument at the next moment; Anatomical structures within potentially occluded areas are hierarchically labeled, and corresponding confidence weights are determined based on the degree of occlusion. The confidence weights are then encoded into transparency parameters, which are then integrated into the real-time 3D spatial data model. The correction module is used to collect real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements, and to dynamically correct the real-time three-dimensional spatial data based on the real-time displacement parameters to maintain spatial synchronization between the imaging data and the anatomical structure. The view adjustment module is used to overlay the corrected three-dimensional anatomical structure model with the surgical field in real time to generate a stereo navigation view, and dynamically adjust the display angle, depth and transparency of the stereo navigation view according to the risk score of the resection location; when the risk score is higher than the preset risk threshold, an ultrasound scan is triggered for secondary calibration and the topology of the obscured area is updated simultaneously. When the risk score is lower than the preset risk threshold, the risk deviation value is output, and the display parameters of the 3D navigation view are fine-tuned based on the risk deviation value.

2. The three-dimensional imaging system for assisting complex chest trauma surgery according to claim 1, characterized in that: After the multimodal image data acquisition is completed, preprocessing is performed simultaneously. The preprocessing steps include: Isotropic sampling processing is performed on CT images to uniformly adjust the slice thickness to a preset standard value; Dynamic range compression and speckle noise suppression are performed on ultrasound images, and speckle noise is eliminated by anisotropic diffusion filtering; CT and ultrasound images were acquired at the end of expiration and peak of inspiration, respectively, and the CT and ultrasound images were fused to correct for respiratory artifacts.

3. The three-dimensional imaging system for assisting complex chest trauma surgery according to claim 1, characterized in that: The model building module constructs a three-dimensional anatomical model containing thymus tissue, vascular networks, and neural distribution through image segmentation processing, including: CT images were processed by image feature comparison and combined with the texture features of ultrasound images for auxiliary segmentation to extract the contour of thymus tissue. By identifying and marking the main blood vessels of a preset diameter, and combining this with ultrasound blood flow imaging to confirm the direction of blood vessel branches, the capillary network can be supplemented. Based on the high-resolution characteristics of ultrasound images, the direction of nerve fibers is identified and a nerve distribution map is generated. The segmentation results from various image sources are spatially matched, and the multimodal image data are fused through respiratory cycle synchronization correction to generate a three-dimensional anatomical model of the thymus region.

4. The three-dimensional imaging system for assisting in complex chest trauma surgery according to claim 1, characterized in that, The step of the surgical field capture module capturing visible light images and near-infrared spectral data of the surgical field in real time and generating real-time three-dimensional spatial data further includes: Dual-channel synchronous imaging is used to acquire high-resolution visible light images and near-infrared spectral information of the surgical field. The dual channels include a visible light channel and a near-infrared channel. The visible light channel is equipped with a polarizing filter, and the near-infrared channel is set with a specific wavelength range to enhance the contrast of vascular imaging. Dynamic threshold segmentation is performed on near-infrared spectral information to extract blood vessel contours, and spatial registration is performed in conjunction with visible light images to align blood vessels with tissue structures.

5. The three-dimensional imaging system for assisting complex chest trauma surgery according to claim 4, characterized in that, The step of the correction module acquiring real-time displacement parameters of the thymus caused by the patient's thoracic respiratory movements includes: By pre-deploying a miniature sensor array in the patient's thoracic cavity, the amplitude and frequency of intercostal muscle contraction can be monitored in real time to obtain thymus displacement data; By selecting respiratory motion feature markers in CT images and combining them with real-time diaphragmatic displacement data from ultrasound images, a three-dimensional spatial displacement vector field is constructed. Collect real-time and historical displacement data of the patient's current respiratory cycle and historical respiratory cycles, perform time series analysis, and output predicted deformation parameters of thymus tissue; The real-time displacement parameters of the thymus are generated by superimposing the three-dimensional spatial displacement vector field with the predicted deformation parameters.

6. The three-dimensional imaging system for assisting complex chest trauma surgery according to claim 5, characterized in that, The correction module dynamically corrects the real-time three-dimensional spatial data based on real-time displacement parameters to maintain spatial synchronization between the image data and the anatomical structure, including the following steps: Based on the three-dimensional spatial displacement vector field and predicted deformation parameters, the coordinate compensation amount of real-time three-dimensional spatial data is calculated, and the image data is adjusted frame by frame according to the coordinate compensation amount. The three-dimensional coordinates of the thoracic anatomical structure are updated at a preset update interval during the respiratory cycle by synchronous acquisition of visible light and near-infrared channels. Based on the real-time monitoring values ​​fed back by the micro-sensor array, the predicted deformation parameters are compared in real time to output the spatial deviation, and the coordinate compensation amount is dynamically corrected according to the spatial deviation. When the spatial deviation exceeds the preset threshold, an alarm mechanism is triggered, and an ultrasound scan is initiated for secondary calibration.

7. The three-dimensional imaging system for assisting complex chest trauma surgery according to claim 6, characterized in that, The step of dynamically adjusting the display angle, depth, and transparency of the stereoscopic navigation view based on the risk score of the cut-off location by the view adjustment module also includes: Calculate the Euclidean distance between the current resection point and the vascular network in the three-dimensional anatomical model, and record it as the first conditional parameter; Collect the minimum distance between the current resection point and the nerve fiber bundle, and record it as the second conditional parameter; The first and second conditional parameters are weighted and fused to generate a risk score. The display angle, depth, and transparency of the 3D navigation view are dynamically adjusted based on the risk score.

8. An electronic device, characterized in that: The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the functions of the complex chest trauma surgery-assisted stereoscopic imaging system as described in any one of claims 1 to 7.

Images