Real-time image guiding and complication early warning system in liver cancer interventional therapy operation

Abstract

The invention belongs to the technical field of image guidance and complication early warning, and discloses a real-time image guidance and complication early warning system in a liver cancer interventional therapy operation. The system comprises a multi-modal image real-time rendering module, an organ deformation modeling module, an intelligent focus recognition display enhancement module, a chemotherapy drug distribution prediction visualization module, a risk relation real-time calculation module and an intelligent early warning decision support module, and obtains a dynamic fusion image report, a deformation parameter report and a focus recognition enhancement report. The method comprises the steps of obtaining a tumor coverage situation report, performing quantitative calculation on a spatial relationship between a treatment area and a dangerous anatomical structure to obtain a risk distance matrix, and performing analysis based on a focus recognition enhancement report and the risk distance matrix to obtain a treatment decision suggestion report. The method has the remarkable advantages of being high in three-dimensional navigation precision degree, good in quantitative prediction effect in the treatment process and large in prospective complication early warning effect.

Description

Technical Field

[0001] This invention relates to the field of image-guided and complication early warning technology, and more specifically, to a real-time image-guided and complication early warning system for interventional treatment of liver cancer. Background Technology

[0002] The real-time image guidance and complication early warning system for interventional treatment of liver cancer is a computer-aided surgical system that integrates advanced image processing, artificial intelligence analysis and real-time data fusion. Its core objective is to enhance the surgeon's perception of lesions and key anatomical structures and to quantitatively predict the treatment process and potential risks, ultimately achieving precise and safe interventional treatment for liver cancer.

[0003] However, traditional real-time image guidance and complication warning systems for interventional liver cancer treatment often suffer from the following shortcomings: First, traditional systems mostly rely on static 3D navigation based on preoperative images, which struggles to address liver deformation and displacement caused by physiological activities such as breathing and heartbeat. This deformation and displacement can lead to discrepancies between the navigation model used by the traditional system and the patient's actual liver position during surgery, potentially resulting in inaccurate lesion localization, incorrect drug injection, and accidental damage to healthy areas. Second, the surgical outcomes of traditional systems largely depend on the personal experience and visual perception of medical personnel, making it difficult for them to identify minute lesions and increasing the risk of tumor recurrence. Third, complication warnings in traditional systems are mostly triggered by monitoring vital signs, with alarms only activated when significant changes occur. By this time, complications have often already formed, failing to effectively mitigate postoperative risks. In summary, effectively addressing the issues of low accuracy in 3D navigation, poor quantitative prediction of the treatment process, and passive complication warnings in traditional systems has become a crucial challenge for current real-time image guidance and complication warning systems for interventional liver cancer treatment.

[0004] In view of this, the present invention proposes a real-time image-guided and complication early warning system for interventional treatment of liver cancer to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution, including: The multimodal image real-time rendering module is used to perform dynamic fusion rendering based on the patient's three-dimensional volume data before surgery, two-dimensional image data during surgery, and deformation parameter reports to obtain a dynamic fusion image report; Furthermore, the steps for dynamic fusion rendering based on preoperative patient 3D volume data, intraoperative 2D image data, and deformation parameter reports include: S1.1: Retrieves three-dimensional volume data and two-dimensional image data from the database, and receives deformation parameter reports output by the organ deformation modeling module in real time; S1.2: Segmentation is performed based on three-dimensional volume data to obtain a three-dimensional model. The three-dimensional model includes the liver outline, vascular tree, tumor region, gallbladder region and stomach region obtained based on the three-dimensional volume data. Among them, the vascular tree includes the hepatic artery, hepatic vein and portal vein. The real-time two-dimensional image data is denoised and real-time feature extraction is performed to obtain a two-dimensional blood vessel feature vector. S1.3: Feature matching is performed between the hepatic artery vascular tree and the two-dimensional vascular feature vector in the 3D model, and the iterative nearest point class algorithm is used to calculate the vascular matrix. The vascular matrix is ​​then input into the 3D model to obtain the first 3D model. The specific calculation formula is as follows: ; Obtain the blood vessel matrix ,in, It is a rigid transformation matrix. For indexing, For the first Blood vessel points on the hepatic artery vascular tree in a 3D model. For the first Blood vessel points on a two-dimensional blood vessel feature vector; S1.4: Input the deformation parameter report into the first 3D model and output the final 3D model; S1.5: Project the final 3D model into a 2D model and fuse the projection result with the 2D image data in step S1.1 to obtain a dynamic fused image report; S1.6: Output the dynamic fusion image report to the intelligent lesion recognition and display enhancement module; The organ deformation modeling module is used to establish and update the patient's personalized respiratory liver motion model and obtain deformation parameter reports based on the patient's real-time respiratory waveform data during surgery. Furthermore, the steps of establishing and updating the patient's personalized respiratory liver motion model and obtaining deformation parameter reports based on the patient's real-time respiratory waveform data during surgery include: S2.1: Collect the patient's respiratory waveform data in real time based on the respiratory equipment, and obtain a three-dimensional model based on step S1.2; S2.2: Identify the respiratory waveform data to obtain the respiratory waveform vector. The identification includes identifying the peak and valley values ​​in the respiratory waveform data, where the peak value is the end of inspiration and the valley value is the end of expiration. S2.3: Based on the linear model of basis functions and constructed according to the respiratory waveform vector, a respiratory liver motion model is obtained. The specific expression of the model is as follows: ; in, For the final 3D model, The respiratory waveform vector The number of basis functions. For the first The weight coefficients of a preset basis function For the first One preset basis function; S2.4: Based on real-time respiratory waveform data, obtain the real-time respiratory waveform vector, input the real-time respiratory waveform vector into the respiratory liver motion model, and output the deformation parameter report; The intelligent lesion recognition and display enhancement module is used to automatically recognize and enhance the display of patient lesions based on dynamic fusion image reports, and obtain a lesion recognition and enhancement report. Furthermore, the steps for automatically identifying and enhancing the display of patient lesions based on dynamic fusion image reports include: S3.1: Retrieve a pre-trained deep learning segmentation model from the database, input the dynamically fused image report into the deep learning segmentation model, and output a probabilistic image report; S3.2: The probabilistic image report is denoised using connected component analysis, and the tumor boundary is optimized using an active contour model to obtain an optimized image report; S3.3: Based on the optimized image report, display enhancement processing is performed to obtain a lesion recognition enhancement report; S3.4: Output the enhanced lesion identification report to the chemotherapy drug distribution prediction visualization module; The chemotherapy drug distribution prediction visualization module is used to simulate and predict the distribution of chemotherapy drugs in the blood vessels of a patient's tumor in real time, and to obtain a tumor coverage report. Furthermore, the steps for real-time simulation and prediction of the distribution of chemotherapy drugs within the patient's tumor blood vessels include: S4.1: Based on the database, retrieve the TACE (transarterial chemoembolization) injection parameters and run the final three-dimensional model to predict the transport and deposition of chemotherapeutic drugs in the tumor vascular network, obtaining a blood flow rate report. The TACE injection parameters include, but are not limited to, injection flow rate, injection volume, and chemotherapeutic drug concentration. The specific calculation formula for the prediction is as follows: ; Get the time point blood flow rate ,in, Initial blood flow velocity, For chemotherapy drug coefficient, For time The cumulative injection volume; S4.2: Using numerical simulation, areas where the blood flow velocity in the blood flow velocity report is lower than the blood flow velocity threshold are marked to obtain a predicted embolization success area report. At the same time, the final concentration of chemotherapy drugs in the patient's blood vessels is calculated to obtain a predicted drug distribution report. S4.3: Based on the enhanced lesion identification report, and by comparing it with the predicted successful embolization area report and the predicted drug distribution report, a tumor coverage report is obtained; S4.4: The predicted successful embolization area report is rendered in real time onto the dynamic fusion image report with a preset color and output to the medical staff receiving end in real time. The predicted drug distribution report is overlaid on the real-time rendered dynamic fusion image report with a different preset color than the predicted successful embolization area report and output to the medical staff receiving end in real time. S4.5: Output the tumor coverage report to the real-time risk relationship calculation module; The real-time risk relationship calculation module is used to quantify the spatial relationship between the treatment area and the dangerous anatomical structures to obtain the risk distance matrix; Furthermore, the steps for quantifying the spatial relationship between the treatment area and the dangerous anatomical structures include: S5.1: Based on step S1.2, obtain the three-dimensional model and extract the key risk structures to obtain the risk grid model; Treatment areas are extracted based on tumor coverage reports and treatment location regions to obtain treatment units. The treatment location regions are acquired in real time through the surgical navigation system. S5.2: Calculate the minimum Euclidean distance between the treatment unit and the risk network model in real time to obtain the risk distance matrix; S5.3: Output the risk distance matrix to the intelligent early warning decision support module; The intelligent early warning decision support module is used to analyze the enhanced lesion identification report and risk distance matrix to obtain a treatment decision recommendation report; Furthermore, the steps for analysis based on lesion identification enhancement reports and risk distance matrices include: S6.1: Retrieve a preset complication risk probability model from the database, input the risk distance matrix into the complication risk probability model, and output a risk probability report; S6.2: Retrieve the complication risk probability rule table from the database and compare it with the risk probability report to generate a risk warning report; S6.3: Retrieve the decision suggestion rule table from the database, and generate a treatment decision suggestion report by comparing it with the risk warning report and the current surgical progress; S6.4: Output risk warning reports and treatment decision recommendation reports to the receiving end of medical personnel; Furthermore, S1: Dynamic fusion rendering is performed based on the patient's three-dimensional volume data before surgery, two-dimensional image data during surgery, and deformation parameter reports to obtain a dynamic fusion image report; S2: Establish and update the patient's personalized respiratory liver motion model, and obtain deformation parameter reports based on the patient's real-time respiratory waveform data during surgery; S3: Based on dynamic fusion image reports, patient lesions are automatically identified and displayed to enhance the results, resulting in a lesion identification and enhancement report; S4: Real-time simulation and prediction of the distribution of chemotherapy drugs in the tumor blood vessels of patients, and a report on tumor coverage; S5: Quantitatively calculate the spatial relationship between the treatment area and the dangerous anatomical structures to obtain the risk distance matrix; S6: Based on the enhanced lesion identification report and risk distance matrix, an analysis is performed to obtain a treatment decision recommendation report.

[0006] The technical effects and advantages of the real-time image guidance and complication early warning system for interventional treatment of liver cancer of this invention are as follows: This invention utilizes dynamic fusion rendering based on preoperative patient 3D volume data, intraoperative 2D image data, and deformation parameter reports to obtain a dynamic fusion image report. It establishes and updates a personalized respiratory liver motion model for the patient and generates a deformation parameter report based on real-time respiratory waveform data during surgery. The system automatically identifies and enhances the display of patient lesions based on the dynamic fusion image report, generating a lesion identification enhancement report. It performs real-time simulation and prediction of the distribution of chemotherapy drugs within the patient's tumor vessels, generating a tumor coverage report. It quantifies the spatial relationship between the treatment area and dangerous anatomical structures, generating a risk distance matrix. Based on the lesion identification enhancement report and the risk distance matrix, it analyzes the data to generate a treatment decision recommendation report. This allows the system to accurately map preoperatively planned vascular and tumor models onto the treatment area through the linkage of a multimodal image real-time rendering module and an organ deformation modeling module. The real-time imaging during surgery significantly improves the accuracy of the 3D navigation model. Furthermore, the invention minimizes the missed detection rate of minute lesions by establishing an intelligent lesion identification and display enhancement module and a chemotherapy drug distribution prediction visualization module. Simultaneously, it provides medical personnel with a forward-looking display of treatment effects, assisting them in adjusting drug injection strategies in real time, thereby minimizing the patient's recurrence risk. Finally, through the collaborative operation of a real-time risk relationship calculation module and an intelligent early warning decision support module, surgical complications are transformed from traditional post-operative responses to pre-operative prevention. This allows medical personnel to provide early warnings before irreversible damage is caused to the patient, thus minimizing the probability of postoperative complications and ensuring postoperative safety. Overall, this invention has significant advantages such as high accuracy of 3D navigation, good quantitative prediction of the treatment process, and a strong forward-looking complication warning function. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the real-time image guidance and complication early warning system for interventional treatment of liver cancer according to the present invention; Figure 2 This is a schematic diagram of the real-time image guidance and complication early warning method for interventional treatment of liver cancer according to the present invention. Detailed Implementation

[0008] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0009] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in the embodiments of this invention are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0010] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0011] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0012] In practice, the server-side equipment deployed in the real-time image guidance and complication early warning system for interventional liver cancer treatment may consist of one or more devices. This system can be implemented as a business instance, a virtual machine, or a hardware device. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node, providing real-time image guidance and complication early warning services to various user terminals. Alternatively, it can be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage each user terminal. Or, it can also be implemented as a server composed of numerous identical or different types of hardware devices, with one or more devices configured to provide real-time image guidance and complication early warning services to various user terminals.

[0013] In terms of implementation, the real-time image guidance and complication early warning system for interventional treatment of liver cancer and the user terminal are mutually compatible. That is, if the real-time image guidance and complication early warning system for interventional treatment of liver cancer is implemented as an application installed on a cloud service platform, then the user terminal is a client that establishes a communication connection with the application; or if the real-time image guidance and complication early warning system for interventional treatment of liver cancer is implemented as a website, then the user terminal is implemented as a webpage; or if the real-time image guidance and complication early warning system for interventional treatment of liver cancer is implemented as a cloud service platform, then the user terminal is implemented as a mini-program in an instant messaging application.

[0014] like Figure 1 The figure shown is a system architecture diagram of a real-time image guidance and complication early warning system for interventional treatment of liver cancer provided in an embodiment of the present invention.

[0015] The real-time image guidance and complication early warning system for interventional treatment of liver cancer described in this invention can be set up in a cloud server. In terms of implementation, it can be implemented as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server or server cluster), or it can be developed as a website. Depending on the functions implemented, the real-time image guidance and complication early warning system for interventional treatment of liver cancer may include a multimodal image real-time rendering module, an organ deformation modeling module, an intelligent lesion recognition and display enhancement module, a chemotherapy drug distribution prediction visualization module, a risk relationship real-time calculation module, and an intelligent early warning decision support module. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can perform a fixed function, stored in the memory of the electronic device.

[0016] In this embodiment of the invention, in the real-time image guidance and complication early warning system for interventional treatment of liver cancer, each of the above-mentioned modules can be implemented independently and can call other modules. Here, "calling" can be understood as one module connecting to multiple modules of another type and providing corresponding services to those connected modules. For example, the intelligent early warning decision support module can call the same information acquisition module to obtain information collected by that module. Based on the above characteristics, in the real-time image guidance and complication early warning system for interventional treatment of liver cancer provided in this embodiment of the invention, without modifying the program code, the applicable scope of the system architecture can be adjusted by adding modules and directly calling them, achieving cluster-based horizontal expansion to quickly and flexibly expand the system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in a virtual device, such as a service instance in a cloud server.

[0017] Please refer to Example 1 Figure 1 As shown in this embodiment, the real-time image-guided and complication early warning system for interventional treatment of liver cancer includes: The multimodal image real-time rendering module is used to perform dynamic fusion rendering based on the patient's three-dimensional volume data before surgery, two-dimensional image data during surgery, and deformation parameter reports to obtain a dynamic fusion image report. Furthermore, the steps for dynamic fusion rendering based on preoperative patient 3D volume data, intraoperative 2D image data, and deformation parameter reports include: S1.1: Retrieves three-dimensional volume data and two-dimensional image data from the database, and receives deformation parameter reports output by the organ deformation modeling module in real time; It should be explained that the three-dimensional volume data was acquired through CT or MRI; the two-dimensional image data was acquired through real-time acquisition of two-dimensional images of digital subtraction angiography during the surgical procedure. S1.2: Segmentation is performed based on three-dimensional volume data to obtain a three-dimensional model. The three-dimensional model includes the liver outline, vascular tree, tumor region, gallbladder region and intestinal region obtained based on the three-dimensional volume data. Among them, the vascular tree includes the hepatic artery, hepatic vein and portal vein. The real-time two-dimensional image data is denoised and real-time feature extraction is performed to obtain a two-dimensional blood vessel feature vector. S1.3: Feature matching is performed between the hepatic artery vascular tree and the two-dimensional vascular feature vector in the 3D model, and the iterative nearest point class algorithm is used to calculate the vascular matrix. The vascular matrix is ​​then input into the 3D model to obtain the first 3D model. The specific calculation formula is as follows: ; Obtain the blood vessel matrix ,in, It is a rigid transformation matrix. For indexing, For the first Blood vessel points on the hepatic artery vascular tree in a 3D model. For the first Blood vessel points on a two-dimensional blood vessel feature vector; It should be explained that rigid transformation matrices include rotations and translations; index To represent the first The nth pair of points, where a pair of points refers to the nth pair of points in the 3D model. Each feature point corresponds to a feature point in the two-dimensional blood vessel feature vector; S1.4: Input the deformation parameter report into the first 3D model and output the final 3D model; S1.5: Project the final 3D model into a 2D model and fuse the projection result with the 2D image data in step S1.1 to obtain a dynamic fused image report; S1.6: Output the dynamic fusion image report to the intelligent lesion recognition and display enhancement module; The organ deformation modeling module is used to establish and update the patient's personalized respiratory liver motion model and obtain a deformation parameter report based on the patient's real-time respiratory waveform data during surgery. Furthermore, the steps of establishing and updating the patient's personalized respiratory liver motion model and obtaining deformation parameter reports based on the patient's real-time respiratory waveform data during surgery include: S2.1: Collect the patient's respiratory waveform data in real time based on the respiratory equipment, and obtain a three-dimensional model based on step S1.2; S2.2: Identify the respiratory waveform data to obtain the respiratory waveform vector. The identification includes identifying the peak and valley values ​​in the respiratory waveform data, where the peak value is the end of inspiration and the valley value is the end of expiration. S2.3: Based on the linear model of basis functions and constructed according to the respiratory waveform vector, a respiratory liver motion model is obtained. The specific expression of the model is as follows: ; in, For the final 3D model, The respiratory waveform vector The number of basis functions. For the first The weight coefficients of a preset basis function For the first One preset basis function; It should be explained that the preset basis functions are used to represent the mode of displacement as a function of phase; S2.4: Based on real-time respiratory waveform data, obtain the real-time respiratory waveform vector, input the real-time respiratory waveform vector into the respiratory liver motion model, and output the deformation parameter report; The intelligent lesion recognition and display enhancement module is used to automatically recognize and enhance the display of patient lesions based on dynamic fusion image reports, and obtain a lesion recognition and enhancement report. Furthermore, the steps for automatically identifying and enhancing the display of patient lesions based on dynamic fusion image reports include: S3.1: Retrieve a pre-trained deep learning segmentation model from the database, input the dynamically fused image report into the deep learning segmentation model, and output a probabilistic image report; It should be explained that the probabilistic image report includes the probability that each pixel in the image identified in each frame of the dynamically fused image report belongs to the background, tumor, microfoci, or invasive margin. S3.2: The probabilistic image report is denoised using connected component analysis, and the tumor boundary is optimized using an active contour model to obtain an optimized image report; S3.3: Based on the optimized image report, display enhancement processing is performed to obtain a lesion recognition enhancement report; It should be explained that enhancement processing refers to, for example, outlining the main tumor area with red lines, highlighting the micro-lesions with yellow highlights, and highlighting the infiltrative edge area with red shading. S3.4: Output the enhanced lesion identification report to the chemotherapy drug distribution prediction visualization module; The chemotherapy drug distribution prediction visualization module is used to simulate and predict the distribution of chemotherapy drugs in the tumor blood vessels of patients in real time and obtain a tumor coverage report. Furthermore, the steps for real-time simulation and prediction of the distribution of chemotherapy drugs within the patient's tumor blood vessels include: S4.1: Based on the database, retrieve the TACE (transarterial chemoembolization) injection parameters and run the final three-dimensional model to predict the transport and deposition of chemotherapeutic drugs in the tumor vascular network, obtaining a blood flow rate report. The TACE injection parameters include, but are not limited to, injection flow rate, injection volume, and chemotherapeutic drug concentration. The specific calculation formula for the prediction is as follows: ; Get the time point blood flow rate ,in, Initial blood flow velocity, For chemotherapy drug coefficient, For time The cumulative injection volume; It should be explained that the chemotherapy drug coefficient is calculated based on the matching degree between the injected chemotherapy drug and the patient's blood vessel diameter; S4.2: Using numerical simulation, areas where the blood flow velocity in the blood flow velocity report is lower than the blood flow velocity threshold are marked to obtain a predicted embolization success area report. At the same time, the final concentration of chemotherapy drugs in the patient's blood vessels is calculated to obtain a predicted drug distribution report. It should be explained that the blood flow rate threshold is manually set and input into the system; S4.3: Based on the enhanced lesion identification report, and by comparing it with the predicted successful embolization area report and the predicted drug distribution report, a tumor coverage report is obtained; It should be explained that the tumor coverage report is used to identify portions of a tumor area that were not expected to be covered. S4.4: The predicted successful embolization area report is rendered in real time onto the dynamic fusion image report with a preset color and output to the medical staff receiving end in real time. The predicted drug distribution report is overlaid on the real-time rendered dynamic fusion image report with a different preset color than the predicted successful embolization area report and output to the medical staff receiving end in real time. S4.5: Output the tumor coverage report to the real-time risk relationship calculation module; The real-time risk relationship calculation module is used to quantify the spatial relationship between the treatment area and the dangerous anatomical structure to obtain a risk distance matrix. Furthermore, the steps for quantifying the spatial relationship between the treatment area and the dangerous anatomical structures include: S5.1: Based on step S1.2, obtain the three-dimensional model and extract the key risk structures to obtain the risk grid model; It should be explained that key risk structures refer to, for example, the gallbladder region and the stomach region; Treatment areas are extracted based on tumor coverage reports and treatment location regions to obtain treatment units. The treatment location regions are acquired in real time through the surgical navigation system. S5.2: Calculate the minimum Euclidean distance between the treatment unit and the risk network model in real time to obtain the risk distance matrix; S5.3: Output the risk distance matrix to the intelligent early warning decision support module; The intelligent early warning decision support module is used to analyze the lesion identification enhancement report and risk distance matrix to obtain a treatment decision suggestion report; Further steps in the analysis based on lesion identification enhancement reports and risk distance matrices include: S6.1: Retrieve a preset complication risk probability model from the database, input the risk distance matrix into the complication risk probability model, and output a risk probability report; It should be explained that complications include, but are not limited to, complications in the gallbladder region, stomach region, and pancreas region; the complication risk probability model means that, for example, when the distance between the treatment unit and the gallbladder region is negative, the probability of gallbladder complications will increase. S6.2: Retrieve the complication risk probability rule table from the database and compare it with the risk probability report to generate a risk warning report; It should be explained that the complication risk probability rule table contains the probability threshold range of all complications. For example, the risk probability threshold range of gallbladder complications is (R1, R2), and there is a gallbladder risk probability value Q1 in the risk probability report. When Q1 is greater than R2, a gallbladder red warning is generated, and when Q1 is less than or equal to R2 and greater than R1, a gallbladder yellow warning is generated. S6.3: Retrieve the decision suggestion rule table from the database, and generate a treatment decision suggestion report by comparing it with the risk warning report and the current surgical progress; It should be explained that step S6.3 means, for example, when the risk warning report is a red warning for the gallbladder and the current surgical procedure is in progress, the treatment decision recommendation report includes stating that there is a risk of injecting therapeutic drugs in the gallbladder area, asking medical personnel to stop the injection operation immediately, and adjusting the TACE hepatic artery chemoembolization injection parameters or finding a supplementary blood supply artery for the tumor. S6.4: Output risk warning reports and treatment decision recommendation reports to the receiving end of medical personnel; In this embodiment, the beneficial effects are achieved through dynamic fusion rendering based on the patient's preoperative 3D volume data, intraoperative 2D image data, and deformation parameter reports, resulting in a dynamic fusion image report. This allows for the establishment and updating of a personalized respiratory liver motion model for the patient. Based on the patient's real-time respiratory waveform data during surgery, a deformation parameter report is obtained. The system automatically identifies and enhances the display of patient lesions based on the dynamic fusion image report, generating a lesion identification enhancement report. Real-time simulation and prediction of the distribution of chemotherapy drugs within the patient's tumor vessels yields a tumor coverage report. The spatial relationship between the treatment area and dangerous anatomical structures is quantified to obtain a risk distance matrix. Analysis based on the lesion identification enhancement report and the risk distance matrix generates a treatment decision recommendation report. This enables the system to accurately render the preoperatively planned vascular and tumor models through the linkage of the multimodal image real-time rendering module and the organ deformation modeling module. Corresponding to real-time images during surgery, this significantly improves the accuracy of the 3D navigation model. Furthermore, the invention minimizes the missed detection rate of minute lesions by establishing an intelligent lesion recognition and display enhancement module and a chemotherapy drug distribution prediction visualization module. Simultaneously, it provides medical personnel with a forward-looking display of treatment effects, assisting them in adjusting drug injection strategies in real time, thereby minimizing the patient's recurrence risk. Finally, through the collaborative operation of a real-time risk relationship calculation module and an intelligent early warning decision support module, surgical complications are transformed from traditional post-operative responses to pre-operative prevention. This allows medical personnel to provide early warnings before irreversible damage is caused to the patient, thus minimizing the probability of postoperative complications and ensuring postoperative safety. Overall, this invention has significant advantages such as high accuracy in 3D navigation, good quantitative prediction of the treatment process, and a strong forward-looking complication warning function.

[0018] Please refer to Example 2 Figure 2 As shown, the parts not described in detail in this embodiment are described in Embodiment 1. A method for real-time image guidance and complication early warning during interventional treatment of liver cancer is provided. The method includes: S1: Dynamic fusion rendering is performed based on the patient's three-dimensional volume data before the operation, the two-dimensional image data during the operation, and the deformation parameter report to obtain a dynamic fusion image report; S2: Establish and update the patient's personalized respiratory liver motion model, and obtain deformation parameter reports based on the patient's real-time respiratory waveform data during surgery; S3: Based on dynamic fusion image reports, patient lesions are automatically identified and displayed to enhance the results, resulting in a lesion identification and enhancement report; S4: Real-time simulation and prediction of the distribution of chemotherapy drugs in the tumor blood vessels of patients, and a report on tumor coverage; S5: Quantitatively calculate the spatial relationship between the treatment area and the dangerous anatomical structures to obtain the risk distance matrix; S6: Based on the enhanced lesion identification report and risk distance matrix, an analysis is performed to obtain a treatment decision recommendation report.

[0019] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the present invention.

Claims

1. A real-time image-guided and complication early warning system for interventional treatment of liver cancer, characterized in that, The system includes: a multimodal image real-time rendering module, an organ deformation modeling module, an intelligent lesion recognition and display enhancement module, a chemotherapy drug distribution prediction and visualization module, a risk relationship real-time calculation module, and an intelligent early warning decision support module, wherein: The multimodal image real-time rendering module is used to perform dynamic fusion rendering based on the patient's three-dimensional volume data before surgery, two-dimensional image data during surgery, and deformation parameter reports to obtain a dynamic fusion image report. The organ deformation modeling module is used to establish and update the patient's personalized respiratory liver motion model and obtain a deformation parameter report based on the patient's real-time respiratory waveform data during surgery. The intelligent lesion recognition and display enhancement module is used to automatically recognize and enhance the display of patient lesions based on dynamic fusion image reports, and obtain a lesion recognition and enhancement report. The chemotherapy drug distribution prediction visualization module is used to simulate and predict the distribution of chemotherapy drugs in the tumor blood vessels of patients in real time and obtain a tumor coverage report. The real-time risk relationship calculation module is used to quantify the spatial relationship between the treatment area and the dangerous anatomical structure to obtain a risk distance matrix. The intelligent early warning decision support module is used to analyze the lesion identification enhancement report and risk distance matrix to obtain a treatment decision suggestion report.

2. The real-time image-guided and complication early warning system for interventional treatment of liver cancer according to claim 1, characterized in that, The steps for dynamic fusion rendering based on preoperative patient 3D volume data, intraoperative 2D image data, and deformation parameter reports include: S1.1: Retrieves three-dimensional volume data and two-dimensional image data from the database, and receives deformation parameter reports output by the organ deformation modeling module in real time; S1.2: Segmentation is performed based on three-dimensional volume data to obtain a three-dimensional model. The three-dimensional model includes the liver outline, vascular tree, tumor region, gallbladder region and stomach region obtained based on the three-dimensional volume data. Among them, the vascular tree includes the hepatic artery, hepatic vein and portal vein. The real-time two-dimensional image data is denoised and real-time feature extraction is performed to obtain a two-dimensional blood vessel feature vector. S1.3: Perform feature matching between the hepatic artery vascular tree and the two-dimensional vascular feature vector in the 3D model, and use the iterative nearest point class algorithm to calculate the vascular matrix. Input the vascular matrix into the 3D model to obtain the first 3D model. S1.4: Input the deformation parameter report into the first 3D model and output the final 3D model; S1.5: Project the final 3D model into a 2D model and fuse the projection result with the 2D image data in step S1.1 to obtain a dynamic fused image report; S1.6: Output the dynamic fusion image report to the intelligent lesion recognition and display enhancement module.

3. The real-time image-guided and complication early warning system for interventional treatment of liver cancer according to claim 2, characterized in that, The steps for establishing and updating a patient's personalized respiratory liver motion model and obtaining a deformation parameter report based on the patient's real-time respiratory waveform data during surgery include: S2.1: Collect the patient's respiratory waveform data in real time based on the respiratory equipment, and obtain a three-dimensional model based on step S1.2; S2.2: Identify the respiratory waveform data to obtain the respiratory waveform vector. The identification includes identifying the peak and valley values ​​in the respiratory waveform data, where the peak value is the end of inspiration and the valley value is the end of expiration. S2.3: Based on the linear model of the basis function and constructed according to the respiratory waveform vector, a respiratory liver motion model is obtained; S2.4: Based on real-time respiratory waveform data, obtain the real-time respiratory waveform vector, input the real-time respiratory waveform vector into the respiratory liver motion model, and output the deformation parameter report.

4. The real-time image-guided and complication early warning system for interventional treatment of liver cancer according to claim 1, characterized in that, The steps for automatic identification and display enhancement of patient lesions based on dynamic fusion image reports include: S3.1: Retrieve a pre-trained deep learning segmentation model from the database, input the dynamically fused image report into the deep learning segmentation model, and output a probabilistic image report; S3.2: The probabilistic image report is denoised using connected component analysis, and the tumor boundary is optimized using an active contour model to obtain an optimized image report; S3.3: Based on the optimized image report, display enhancement processing is performed to obtain a lesion recognition enhancement report; S3.4: Output the enhanced lesion identification report to the chemotherapy drug distribution prediction visualization module.

5. The real-time image-guided and complication early warning system for interventional treatment of liver cancer according to claim 1, characterized in that, The steps for real-time simulation and prediction of the distribution of chemotherapy drugs within the tumor blood vessels of patients include: S4.1: Based on the database, retrieve the TACE hepatic artery chemoembolization injection parameters and run the final three-dimensional model to predict the transport and deposition of chemotherapeutic drugs in the tumor vascular network and obtain a blood flow rate report. The hepatic artery chemoembolization injection parameters include, but are not limited to, injection flow rate, injection volume and chemotherapeutic drug concentration. S4.2: Using numerical simulation, areas where the blood flow velocity in the blood flow velocity report is lower than the blood flow velocity threshold are marked to obtain a predicted embolization success area report. At the same time, the final concentration of chemotherapy drugs in the patient's blood vessels is calculated to obtain a predicted drug distribution report. S4.3: Based on the enhanced lesion identification report, and by comparing it with the predicted successful embolization area report and the predicted drug distribution report, a tumor coverage report is obtained; S4.4: The predicted successful embolization area report is rendered in real time onto the dynamic fusion image report with a preset color and output to the medical staff receiving end in real time. The predicted drug distribution report is overlaid on the real-time rendered dynamic fusion image report with a different preset color than the predicted successful embolization area report and output to the medical staff receiving end in real time. S4.5: Output the tumor coverage report to the real-time risk relationship calculation module.

6. The real-time image guidance and complication early warning system for interventional treatment of liver cancer according to claim 1, characterized in that, The steps for quantifying the spatial relationship between the treatment area and the dangerous anatomical structures include: S5.1: Based on step S1.2, obtain the three-dimensional model and extract the key risk structures to obtain the risk grid model; Treatment areas are extracted based on tumor coverage reports and treatment location regions to obtain treatment units. The treatment location regions are acquired in real time through the surgical navigation system. S5.2: Calculate the minimum Euclidean distance between the treatment unit and the risk network model in real time to obtain the risk distance matrix; S5.3: Output the risk distance matrix to the intelligent early warning decision support module.

7. The real-time image-guided and complication early warning system for interventional treatment of liver cancer according to claim 1, characterized in that, The steps involved in the analysis based on lesion identification enhancement reports and risk distance matrices include: S6.1: Retrieve a preset complication risk probability model from the database, input the risk distance matrix into the complication risk probability model, and output a risk probability report; S6.2: Retrieve the complication risk probability rule table from the database and compare it with the risk probability report to generate a risk warning report; S6.3: Retrieve the decision suggestion rule table from the database, and generate a treatment decision suggestion report by comparing it with the risk warning report and the current surgical progress; S6.4: Output risk warning reports and treatment decision recommendation reports to the receiving end of medical personnel.

8. A method for real-time image guidance and complication early warning during interventional therapy for liver cancer, implemented according to any one of claims 1-7, characterized in that, The work includes the following steps: S1: Dynamic fusion rendering is performed based on the patient's three-dimensional volume data before surgery, two-dimensional image data during surgery, and deformation parameter reports to obtain a dynamic fusion image report; S2: Establish and update the patient's personalized respiratory liver motion model, and obtain deformation parameter reports based on the patient's real-time respiratory waveform data during surgery; S3: Based on dynamic fusion image reports, patient lesions are automatically identified and displayed to enhance the results, resulting in a lesion identification and enhancement report; S4: Real-time simulation and prediction of the distribution of chemotherapy drugs in the tumor blood vessels of patients, and a report on tumor coverage; S5: Quantitatively calculate the spatial relationship between the treatment area and the dangerous anatomical structures to obtain the risk distance matrix; S6: Based on the enhanced lesion identification report and risk distance matrix, an analysis is performed to obtain a treatment decision recommendation report.

Images