Medical evaluation method and device, electronic equipment and storage medium

Abstract

The application discloses a medical evaluation method and device, electronic equipment and a storage medium. The medical evaluation method comprises the following steps: acquiring a target medical image of multiple sequences; performing target segmentation on the target medical image of multiple sequences to determine at least one of a dynamic and static vein tissue segmentation result and a vein abnormal tissue segmentation result; performing vein abnormality identification based on at least one of the dynamic and static vein tissue segmentation result and the vein abnormal tissue segmentation result to obtain a vein abnormality identification result; and performing inferior vena cava occlusion evaluation based on the vein abnormality identification result to obtain an evaluation result. The method can automatically segment the dynamic and static vein tissue of the target medical image of multiple sequences, accurately detect the lesion tissue based on the identified vein vessel position, comprehensively evaluate the inferior vena cava occlusion surgery, and effectively reduce the diagnosis and treatment time, the surgical risk and the labor cost.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and in particular to a medical assessment method, device, electronic device, and storage medium. Background Technology

[0002] With the rapid increase in the incidence of thrombotic and cardiovascular diseases, the incidence of pulmonary embolism is also rising. Inferior vena cava occlusion can effectively reduce the risk of pulmonary embolism; however, the procedure is highly complex and requires precise and comprehensive preoperative and postoperative assessments to deliver the occluder via catheter to the appropriate location in the inferior vena cava and ensure the best surgical outcome.

[0003] Existing technologies extract arterial or venous tissue for specific tissue detection using single-phase images, which are easily affected by factors such as scanning timing and contrast agents, leading to inaccurate detection results and poor surgical evaluation outcomes. Summary of the Invention

[0004] In view of the above problems, this application provides a medical assessment method, device, electronic device and storage medium that can perform fully automatic arterial and venous tissue segmentation on multi-sequence target medical images and accurately detect lesions based on the identified venous vessel locations to conduct a comprehensive assessment of inferior vena cava occlusion surgery, effectively reducing treatment time, surgical risks and labor costs.

[0005] In a first aspect, this application provides a medical assessment method, which includes: acquiring a multi-sequence target medical image; performing target segmentation on the multi-sequence target medical image to obtain at least one of a segmentation result of arteriovenous tissue and a segmentation result of abnormal venous tissue; performing venous abnormality identification based on at least one of the segmentation results of arteriovenous tissue and abnormal venous tissue to obtain a venous abnormality identification result; and performing inferior vena cava occlusion assessment based on the venous abnormality identification result to obtain an assessment result.

[0006] In the technical solution of this application embodiment, firstly, a multi-sequence target medical image is acquired, then the multi-sequence target medical image is segmented to obtain at least one of the determined arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result. Then, based on at least one of the arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result, venous abnormality identification is performed to obtain the venous abnormality identification result. Finally, based on the venous abnormality identification result, the inferior vena cava occlusion assessment is performed to obtain the assessment result. This method can perform fully automatic arteriovenous tissue segmentation on multi-sequence target medical images and accurately detect lesions based on the identified venous vessel locations to conduct a comprehensive assessment of inferior vena cava occlusion surgery, effectively reducing treatment time, surgical risks, and labor costs.

[0007] In some embodiments, the vein anomaly identification results include vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results. Based on the vein anomaly identification results, an inferior vena cava occlusion assessment is performed to obtain an assessment result, including at least one of the following: evaluating the effective path for inferior vena cava occlusion based on at least one of the vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results to obtain a target occlusion path as the assessment result; and evaluating the safe area for release of the inferior vena cava occluder based on the vein thrombosis identification results to obtain a target occlusion area as the assessment result.

[0008] In some embodiments, the arteriovenous tissue segmentation results include arterial tissue segmentation results and venous tissue segmentation results, and the venous abnormal tissue segmentation results include thrombus tissue segmentation results and calcified tissue segmentation results; based on at least one of the arteriovenous tissue segmentation results and venous abnormal tissue segmentation results, venous abnormality identification is performed to obtain venous abnormality identification results, including at least one of the following: based on the venous tissue segmentation results and thrombus tissue segmentation results, venous thrombosis identification is performed to obtain venous thrombosis identification results; based on the venous tissue segmentation results and calcified tissue segmentation results, venous calcification identification is performed to obtain venous calcification identification results; based on the arterial tissue segmentation results and venous tissue segmentation results, venous stenosis identification is performed to obtain venous stenosis identification results.

[0009] In some embodiments, the target medical image of the multi-sequence includes plain scan image, arterial phase image, and venous phase image; based on the venous tissue segmentation result and the thrombus tissue segmentation result, venous thrombosis identification is performed to obtain venous thrombosis identification result, including: classifying the venous tissue of the target medical image of the multi-sequence according to the venous tissue segmentation result and the venous phase image to obtain venous tissue classification result; and identifying the venous thrombosis site of the target medical image of the multi-sequence according to the venous tissue classification result and the thrombus tissue segmentation result to obtain venous thrombosis identification result.

[0010] In some embodiments, vein calcification is identified based on vein tissue segmentation results and calcified tissue segmentation results to obtain vein calcification identification results, including: identifying vein calcification sites in multi-sequence target medical images based on vein tissue classification results and calcified tissue segmentation results to obtain vein calcification identification results.

[0011] In some embodiments, based on the arterial tissue segmentation results and the venous tissue segmentation results, venous stenosis identification is performed to obtain venous stenosis identification results, including: merging the arterial tissue segmentation results and the venous tissue segmentation results according to the positional relationship between the arterial tissue and the venous tissue in the multi-sequence target medical image to obtain a first venous tissue separation result, and merging the arterial tissue segmentation results and the venous tissue classification results to obtain a second venous tissue separation result; extracting and identifying the lumen centerline of the venous tissue using image processing technology based on the first venous tissue separation result and the second venous tissue separation result to obtain a lumen centerline identification result; and calculating the venous lumen morphological parameters based on the lumen centerline identification result to obtain venous stenosis identification results, wherein the venous lumen morphological parameters include at least one of diameter, area, mean, variance, and stenosis rate.

[0012] In some embodiments, target segmentation is performed on the multi-sequence target medical image to determine at least one of arterial and venous tissue segmentation results and abnormal venous tissue segmentation results, including: performing arterial tissue segmentation on the multi-sequence target medical image based on the arterial phase image to obtain arterial tissue segmentation results, and performing venous tissue segmentation on the multi-sequence target medical image based on the venous phase image to obtain venous tissue segmentation results; and performing venous thrombosis segmentation and venous calcification segmentation on the multi-sequence target medical image based on the plain scan image and the venous tissue segmentation results to obtain thrombosis tissue segmentation results and calcification tissue segmentation results.

[0013] In some embodiments, after identifying vein abnormalities based on at least one of the arteriovenous tissue segmentation results and the vein abnormality tissue segmentation results, and obtaining the vein abnormality identification results, the method further includes: calculating the morphological parameters of the vein abnormality tissue based on the vein thrombosis identification results and the vein calcification identification results, wherein the morphological parameters of the vein abnormality tissue include at least one of coordinates, length, volume, mean, and volume percentage.

[0014] In some embodiments, the inferior vena cava occlusion assessment is performed based on the vein anomaly identification results to obtain the assessment results. The assessment also includes: evaluating the size of the inferior vena cava occluder based on the vein lumen morphology parameters to obtain the target size as the assessment result.

[0015] In some embodiments, the method further includes generating a structured report based on at least two of the following: arteriovenous tissue segmentation results, abnormal venous tissue segmentation results, abnormal venous identification results, venous tissue classification results, evaluation results, venous lumen morphological parameters, and abnormal venous tissue morphological parameters.

[0016] On the other hand, this application provides a medical assessment device, comprising: an acquisition module for acquiring multi-sequence target medical images; a segmentation module for segmenting the multi-sequence target medical images to determine at least one of arteriovenous tissue segmentation results and abnormal venous tissue segmentation results; an identification module for identifying venous abnormalities based on at least one of the arteriovenous tissue segmentation results and abnormal venous tissue segmentation results to obtain a venous abnormality identification result; and an assessment module for assessing inferior vena cava occlusion based on the venous abnormality identification result to obtain an assessment result.

[0017] On the other hand, this application provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method of any of the above embodiments.

[0018] On the other hand, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method of any of the above embodiments.

[0019] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0020] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a medical assessment method according to an embodiment of this application is shown; Figure 2 A schematic diagram of arteriovenous segmentation of the lower limbs according to an embodiment of this application is shown; Figure 3 A schematic diagram showing the vein morphology parameter results of an embodiment of this application is shown; Figure 4 This illustration shows a schematic diagram of the results of the inferior vena cava occlusion risk assessment according to an embodiment of this application; Figure 5 A block diagram of a medical assessment device according to an embodiment of this application is shown; Figure 6 A schematic diagram of an electronic device according to an embodiment of this application is shown. Detailed Implementation

[0021] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0023] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0024] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0025] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0026] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0027] In the description of the embodiments of this application, the technical terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of this application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0028] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0029] The clinical manifestations of pulmonary embolism include sudden chest pain, chest tightness, dyspnea and cyanosis. Severe patients may develop shock. With the rapid increase in the incidence of thrombotic and cardiovascular diseases, the incidence of pulmonary embolism is also rising.

[0030] The vast majority of pulmonary emboli originate from the inferior vena cava system (such as the deep veins of the lower extremities and the pelvic venous plexus). When these thrombi break off and travel into the pulmonary artery with the bloodstream, they can cause severe respiratory and circulatory disturbances, and even be life-threatening. For patients at high risk of thrombosis (such as those who are bedridden for extended periods, have undergone surgery, have tumors, or have contraindications to anticoagulation) and who are at risk of thrombus detachment, inferior vena cava occlusion can effectively reduce the risk of pulmonary embolism.

[0031] Inferior vena cava occlusion is a highly complex procedure requiring precise and comprehensive preoperative assessment. Optimal selection of the puncture approach and placement of the occluder are crucial for surgical success. First, a safe puncture approach must be determined. Preoperative assessment confirms the anatomical abnormalities of the iliac vein and lower extremity renal veins, including the presence of thrombi and calcification. Second, the diameter of the inferior vena cava must be determined to select the appropriate occluder size, assess its course for any abnormalities, and identify the location of the renal vein orifice to determine the safe release area for the occluder. Based on this preoperative assessment information, a preoperative plan is developed to guide the intraoperative delivery of the occluder via catheter to the appropriate location within the inferior vena cava (generally below the renal vein orifice). After release, the occluder is fixed within the vessel, forming a "filter" to intercept thrombi. Simultaneously, the venous patency after inferior vena cava occlusion needs to be assessed to evaluate postoperative efficacy.

[0032] Some implementations segment veins based on the original enhanced image of the veins, compare the veins with a standard atlas matched to the current patient, and obtain the vascular detection results.

[0033] Some embodiments achieve non-invasive detection of the hardness of lower extremity venous thrombosis by comprehensively analyzing magnetic resonance and ultrasound images of the target object. Based on bilateral lower extremity full-length localization images and historical thrombosis localization images, multimodal thrombosis information of each lower extremity deep vein thrombosis in the target object is determined. This multimodal thrombosis information includes thrombosis type and thrombosis properties, enabling the acquisition of detailed information about the thrombosis from different perspectives. This detailed information can comprehensively reflect the current state of the thrombosis, thereby improving the accuracy of hardness measurement of lower extremity venous thrombosis.

[0034] Some embodiments analyze target vein images using a trained venous thrombosis detection model to obtain venous thrombosis detection results.

[0035] However, the above methods generally have some problems. On the one hand, because the veins in the lower limbs are long and run parallel to the arteries in the lower limbs, and there is venous return, they are greatly affected by the timing of the scan and the flow rate of the contrast agent. Therefore, the arterial phase image may have venous contrast, and the venous phase image may have arterial contrast. So it is difficult to extract clean arterial or venous tissue from a single phase. On the other hand, it is difficult to establish a database based on thrombosis detection in actual clinical practice, and its practicality is not strong.

[0036] In view of this, this application proposes a medical evaluation method, device, electronic device, and storage medium that can perform fully automated arteriovenous tissue segmentation on multi-sequence target medical images and accurately detect lesions based on the identified venous vessel locations for a comprehensive evaluation of inferior vena cava occlusion surgery. The medical evaluation method of this application obtains evaluation results by performing relevant processing operations on medical images. The evaluation results can be used for surgical reference or to assist doctors in making customized treatment plans, effectively reducing diagnosis and treatment time, surgical risks, and labor costs.

[0037] In the technical solution of this application embodiment, firstly, a multi-sequence target medical image is acquired, then the multi-sequence target medical image is segmented to obtain at least one of the determined arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result. Then, based on at least one of the arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result, venous abnormality identification is performed to obtain the venous abnormality identification result. Finally, based on the venous abnormality identification result, the inferior vena cava occlusion assessment is performed to obtain the assessment result. This method can perform fully automatic arteriovenous tissue segmentation on multi-sequence target medical images and accurately detect lesions based on the identified venous vessel locations to conduct a comprehensive assessment of inferior vena cava occlusion surgery, effectively reducing treatment time, surgical risks, and labor costs.

[0038] Figure 1 A flowchart of a medical assessment method according to an embodiment of this application is shown.

[0039] like Figure 1 As shown, the medical assessment method 100 provided in this application includes steps S110 to S140.

[0040] Step S110: Obtain the target medical image with multiple sequences.

[0041] For example, the target medical images of multiple sequences may include multiple sequences of medical images that have undergone multiple tissue overlays on the target site. The target site may be, for example, a body part containing relevant tissues. For example, the target medical images of multiple sequences for evaluating inferior vena cava occlusion surgery may be multiple sequence CT images covering the chest, abdomen and lower limbs that contain human arterial and venous vascular tissues and related abnormal tissues. The multiple sequence images may include plain scan images, arterial phase images and venous phase images.

[0042] Step S120: Perform target segmentation on the multi-sequence target medical image to determine at least one of the segmentation results of arteriovenous tissue and abnormal venous tissue.

[0043] For example, the segmentation results of arterial and venous tissues can include segmentation results of arterial tissues and segmentation results of venous tissues. The segmentation results of abnormal venous tissues can include segmentation results of lesions in the veins, such as thrombi or calcifications. Target segmentation can be performed on registered multi-sequence CT images (multi-sequence target medical images) to obtain segmentation results of arterial tissues containing arterial tissue segmentation labels, or segmentation results of venous tissues containing venous tissue segmentation labels, or segmentation results of abnormal venous tissues containing thrombus tissue labels and calcification tissue labels. Target segmentation can be achieved through a pre-designed and trained neural network model. For example, arterial tissues in the arterial phase images, venous tissues in the venous phase images, and thrombi and calcifications in the plain scan images of some acquired multi-sequence CT images of the chest, abdomen, and lower limbs can be labeled, and the segmentation model (neural network model) can be trained based on the dataset images containing the labels of each tissue to obtain a trained model. For details, please refer to the description of the model training section below.

[0044] Step S130: Based on at least one of the arteriovenous tissue segmentation results and the abnormal vein tissue segmentation results, perform vein abnormality identification to obtain vein abnormality identification results.

[0045] For example, the vein anomaly identification result can be the identification result of the specific location or extent of the lesion tissue in the vein, such as identifying whether a thrombus exists. If it exists, the location and extent of the thrombus can be further confirmed (e.g., located in a branch of a blood vessel such as the left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, or left / right superficial femoral vein), or the location and extent of calcification, or the location of venous stenosis. The vein tissue in multi-sequence CT images can be separated based on the arteriovenous tissue segmentation result. Based on the vein abnormality tissue segmentation result, it can be confirmed whether there is lesion tissue such as thrombus, calcification, or stenosis in the separated clean vein. Thus, based on the arteriovenous tissue segmentation result or the vein abnormality tissue segmentation result, the specific location of the abnormal tissue can be confirmed, and the vein anomaly identification result can be obtained.

[0046] Step S140: Based on the results of vein abnormality identification, perform inferior vena cava occlusion assessment to obtain the assessment results.

[0047] For example, the evaluation results may include preoperative or postoperative assessments related to the inferior vena cava occlusion procedure. For instance, the inferior vena cava occlusion surgical approach may be evaluated based on the identification of stenosis / calcification / thrombosis in the vein, or the selection and fixation area of ​​the occluder for the inferior vena cava occlusion procedure may be evaluated. The evaluation results may include the effective path of the surgical approach, the applicable size of the occluder, the safe area for occluder release, etc.

[0048] In the technical solution of this application embodiment, firstly, a multi-sequence target medical image is acquired, then the multi-sequence target medical image is segmented to obtain at least one of the determined arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result. Then, based on at least one of the arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result, venous abnormality identification is performed to obtain the venous abnormality identification result. Finally, based on the venous abnormality identification result, the inferior vena cava occlusion assessment is performed to obtain the assessment result. This method can perform fully automatic arteriovenous tissue segmentation on multi-sequence target medical images and accurately detect lesions based on the identified venous vessel locations to conduct a comprehensive assessment of inferior vena cava occlusion surgery, effectively reducing treatment time, surgical risks, and labor costs.

[0049] In one example, the medical evaluation method of this application is implemented based on a pre-trained network model framework. This model is capable of fully automatic arterial and venous segmentation and separation of multi-sequence target medical images, and identification of lesion tissue, thereby providing a comprehensive evaluation of inferior vena cava occlusion surgery. The training process of the model is described in detail below.

[0050] Step 1: Dataset Construction First, collect multi-sequence CT images, including plain scan, arterial phase, and venous phase, covering the chest, abdomen, and lower extremities. The reconstructed image matrix should be greater than or equal to 512. 512. The dataset underwent rigorous screening and preprocessing to remove data that did not meet the inclusion criteria, ensuring image quality. Then, the multi-sequence CT images were registered using rigid or non-rigid registration algorithms to align the arterial and venous phase images to the plain scan images, generating registered arterial and venous phase images. The dataset was then expanded using techniques such as translation, rotation, and scaling to improve the robustness of the algorithm. Finally, the dataset was divided proportionally into training, optimization, and test sets.

[0051] Step 2: Data Labeling Accurate manual annotation is performed on the training set images in the dataset. On the registered arterial phase images, labels are added for at least the aorta, left / right internal iliac arteries, left / right external iliac arteries, left / right common femoral arteries, left / right deep femoral arteries, and left / right superficial femoral arteries. On the registered venous phase images, labels are added for at least the inferior vena cava, left / right renal veins, left / right external iliac veins, left / right common femoral veins, left / right deep femoral veins, and left / right superficial femoral veins. On the plain scan images, labels are added for at least thrombi and calcifications. Different tissues are labeled with different labels. This annotation work should be performed by physicians or experts with professional knowledge and experience to ensure accuracy and consistency.

[0052] Step 3: Model Training The registered arterial phase image is designated as the first input image. An image with a single label containing at least the aorta, left / right internal iliac artery, left / right external iliac artery, left / right common femoral artery, left / right deep femoral artery, and left / right superficial femoral artery is designated as the first input mask image. The first input image and the first input mask image are used as inputs to the training set. A neural network is designed to learn features and train to generate the first segmentation sub-model. An image with multiple tissue labels containing at least the aorta, left / right internal iliac artery, left / right external iliac artery, left / right common femoral artery, left / right deep femoral artery, and left / right superficial femoral artery is designated as the second input mask image. The first input image, the first input mask image, and the second input mask image are used as inputs to the training set. A convolutional neural network model that simultaneously handles multiple segmentation tasks and combines a hybrid attention mechanism is designed to learn features and train to generate the first classification sub-model.

[0053] The registered venous phase image is designated as the second input image. An image with a single label containing at least the inferior vena cava, left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, and left / right superficial femoral vein is designated as the third input mask image. The second and third input mask images are used as inputs to the training set. A neural network is designed to learn features and train to generate the second segmentation sub-model. An image with multiple tissue labels containing at least the inferior vena cava, left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, and left / right superficial femoral vein is designated as the fourth input mask image. The second, third, and fourth input mask images are used as inputs to the training set. A convolutional neural network model that simultaneously handles multiple segmentation tasks and incorporates a hybrid attention mechanism is designed to learn features and train to generate the second classification sub-model.

[0054] The plain scan image is designated as the third input image, and the image containing the thrombus label is designated as the fifth input mask image. The third input image and the fifth input mask image are used as inputs to the training set. A neural network is designed to learn features and train to generate the third segmentation sub-model.

[0055] The flat scan image is designated as the third input image, and the image containing calcification labels is designated as the sixth input mask image. The third input image and the sixth input mask image are used as inputs to the training set. A neural network is designed to learn features and train to generate the fourth segmentation sub-model.

[0056] Finally, the first, second, third, and fourth segmentation sub-models and the first and second classification sub-models obtained above are optimized using a validation set to obtain the model framework for medical assessment.

[0057] Next, based on the trained medical assessment model framework, target segmentation and target recognition are performed on the multi-sequence CT images (multi-sequence target medical images) of the test set to assess inferior vena cava occlusion.

[0058] For example, target segmentation is performed on a multi-sequence target medical image to determine at least one of the arterial and venous tissue segmentation results and venous abnormal tissue segmentation results. For example, firstly, arterial tissue segmentation is performed on the multi-sequence target medical image based on the arterial phase image to obtain the arterial tissue segmentation result, and venous tissue segmentation is performed on the multi-sequence target medical image based on the venous phase image to obtain the venous tissue segmentation result; then, based on the plain scan image and the venous tissue segmentation result, venous thrombosis segmentation and venous calcification segmentation are performed on the multi-sequence target medical image respectively to obtain thrombosis tissue segmentation results and calcification tissue segmentation results.

[0059] Specifically, the target medical images of multiple sequences may include plain scan images, arterial phase images, and venous phase images. First, the arterial phase images and venous phase images are aligned to the plain scan images using a registration method to generate registered arterial phase images and venous phase images. The arterial phase images and venous phase images mentioned in subsequent steps are all registered images, which will not be elaborated here.

[0060] Then, the registered arterial phase image is input into the first segmentation sub-model to obtain the first tissue segmentation result (arterial tissue segmentation result), and obtain tissue segmentation labels that include at least the aorta, left / right internal iliac artery, left / right external iliac artery, left / right common femoral artery, left / right deep femoral artery, and left / right superficial femoral artery.

[0061] The registered venous phase image is input into the second segmentation sub-model to obtain the second tissue segmentation result (venous tissue segmentation result). The plain scan image and the second tissue segmentation result are input into the third segmentation sub-model to obtain the third tissue segmentation result of the thrombus (thrombus tissue segmentation result), which is the thrombus segmentation image of the venous tissue. The plain scan image and the second tissue segmentation result are input into the fourth segmentation sub-model to obtain the fourth tissue segmentation result of calcification (calcified tissue segmentation result), which is the calcified segmentation image of the venous tissue.

[0062] In the technical solution of this application embodiment, firstly, arterial tissue segmentation is performed on the target medical image of multiple sequences based on the arterial phase image to obtain the arterial tissue segmentation result; and secondly, venous tissue segmentation is performed on the target medical image of multiple sequences based on the venous phase image to obtain the venous tissue segmentation result. Then, based on the plain scan image and the venous tissue segmentation result, venous thrombosis segmentation and venous calcification segmentation are performed on the target medical image of multiple sequences to obtain thrombosis tissue segmentation result and calcification tissue segmentation result. By combining image processing and deep learning technology, arterial and venous tissues and lesions in multi-sequence CT images are automatically segmented for inferior vena cava occlusion assessment, which greatly reduces the diagnosis and treatment time and eliminates the need for manual operation.

[0063] For example, the arteriovenous tissue segmentation results include arterial tissue segmentation results and venous tissue segmentation results, and the venous abnormal tissue segmentation results include thrombus tissue segmentation results and calcified tissue segmentation results; based on at least one of the arteriovenous tissue segmentation results and venous abnormal tissue segmentation results, venous abnormality identification is performed to obtain venous abnormality identification results, including at least one of the following: for example, based on the venous tissue segmentation results and thrombus tissue segmentation results, venous thrombosis identification is performed to obtain venous thrombosis identification results; for another example, based on the venous tissue segmentation results and calcified tissue segmentation results, venous calcification identification is performed to obtain venous calcification identification results; for yet another example, based on the arterial tissue segmentation results and venous tissue segmentation results, venous stenosis identification is performed to obtain venous stenosis identification results.

[0064] For example, the target medical image of the multi-sequence includes plain scan image, arterial phase image, and venous phase image; based on the vein tissue segmentation result and the thrombus tissue segmentation result, venous thrombosis identification is performed to obtain the venous thrombosis identification result. For example, firstly, based on the vein tissue segmentation result and the venous phase image, the vein tissue of the target medical image of the multi-sequence is classified to obtain the vein tissue classification result; then, based on the vein tissue classification result and the thrombus tissue segmentation result, the venous thrombosis site of the target medical image of the multi-sequence is identified to obtain the venous thrombosis identification result.

[0065] Specifically, the registered venous phase image and the second tissue segmentation result are first input into the second classification sub-model to obtain the second tissue classification result (venous tissue classification result). The second tissue classification result includes at least the multi-tissue segmentation labels of the inferior vena cava, left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, and left / right superficial femoral vein.

[0066] Then, based on the third tissue segmentation results and the second classification results, it is confirmed whether there is a thrombus in the vein. If so, the location and extent of the thrombus are further confirmed, such as whether it belongs to a branch of a blood vessel such as the left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, or left / right superficial femoral vein, thereby obtaining the venous thrombus identification result.

[0067] In the technical solution of this application embodiment, firstly, based on the venous tissue segmentation results and venous phase images, the target medical images of multiple sequences are classified into venous tissues to obtain venous tissue classification results. Then, based on the venous tissue classification results and thrombus tissue segmentation results, the target medical images of multiple sequences are identified into venous thrombus sites to obtain venous thrombus identification results. By combining image processing with deep learning technology, the specific location of venous thrombus in multi-sequence CT images is identified to perform inferior vena cava occlusion assessment, thereby improving the accuracy of the assessment results.

[0068] For example, based on the segmentation results of vein tissue and calcified tissue, vein calcification is identified to obtain vein calcification identification results. For instance, based on the classification results of vein tissue and the segmentation results of calcified tissue, vein calcification sites are identified in multi-sequence target medical images to obtain vein calcification identification results.

[0069] Specifically, based on the results of the fourth tissue segmentation and the second classification, it is confirmed whether calcification exists in the vein. If it exists, the location and extent of the calcification are further confirmed, such as whether it belongs to a branch of a blood vessel, such as the left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, or left / right superficial femoral vein, thereby obtaining the venous calcification identification result.

[0070] In the technical solution of this application embodiment, based on the classification results of vein tissue and the segmentation results of calcified tissue, the vein calcification sites are identified in the multi-sequence target medical images to obtain the vein calcification identification results. By combining image processing and deep learning technology, the specific locations where vein calcification exists in the multi-sequence CT images are identified to perform inferior vena cava occlusion assessment, thereby improving the accuracy of the assessment results.

[0071] For example, based on the arterial tissue segmentation results and the venous tissue segmentation results, venous stenosis identification is performed to obtain venous stenosis identification results. For instance, firstly, according to the positional relationship between arterial tissue and venous tissue in the target medical image of multiple sequences, the arterial tissue segmentation results and the venous tissue segmentation results are merged to obtain a first venous tissue separation result, and the arterial tissue segmentation results and the venous tissue classification results are merged to obtain a second venous tissue separation result; then, based on the first venous tissue separation result and the second venous tissue separation result, the venous lumen centerline is extracted and identified using image processing technology to obtain the venous lumen centerline identification result; then, based on the venous lumen centerline identification result, the venous lumen morphological parameters are calculated to obtain the venous stenosis identification result, wherein the venous lumen morphological parameters include at least one of diameter, area, mean, variance, and stenosis rate.

[0072] Specifically, the first and second tissue segmentation results are first merged based on the positional relationship (e.g., pixel coordinates) of arteries and veins in the image to obtain the merged fifth tissue segmentation result (first vein separation result). The merging principle uses the highest priority segmentation result as the final display result of the fifth tissue segmentation, that is, the key results (e.g., veins) are displayed. Then, the first and second tissue segmentation results are merged based on the image positional relationship to obtain the merged sixth tissue segmentation result (second vein separation result). The merging principle uses the highest priority segmentation result (e.g., different parts of the vein) as the final display result of the sixth tissue segmentation. By merging the first and second segmentation results, effective separation of arteries and veins can be achieved, eliminating arterial interference.

[0073] Alternatively, the first tissue classification result of the artery can be obtained based on the first sub-model, and then merged with the second tissue classification result to serve as an auxiliary reference result for venous tissue separation.

[0074] Then, based on the fifth tissue segmentation result, traditional image processing techniques, such as thinning methods, are used to extract the lumen centerline of the venous tissue and establish a tree-shaped lumen structure. Based on the sixth segmentation result, the obtained lumen centerlines are identified and classified to obtain the lumen centerlines of the inferior vena cava, left / right renal vein, left / right external iliac vein, left / right common femoral vein, left / right deep femoral vein, and left / right superficial femoral vein, and the veins are classified and named to obtain the third tissue classification result (lumen centerline identification result).

[0075] Next, the cross-sectional parameters (venous lumen morphology parameters) of the vein at its location are calculated along the centerline of the vein lumen, such as diameter, area, mean, variance, and stenosis rate. When the diameter of the lumen at the identified site differs significantly from the diameter of the lumen at surrounding sites, for example, if the diameter of the lumen at the current location is less than a preset threshold, it indicates that there is a stenotic lesion at that site, thus obtaining the result of vein stenosis identification.

[0076] It should be noted that this method supports volume rendering, surface rendering, volume rendering, MIP (maximum density projection), MinIP (minimum density projection), or tissue network model rendering in any combination of tissues, such as veins, arteries, thrombi, or calcifications obtained through target segmentation and target classification. It also supports the adjustment or combination of any tissue color or protocol for display.

[0077] In the technical solution of this application embodiment, firstly, based on the positional relationship between arterial and venous tissues in multi-sequence target medical images, the arterial tissue segmentation results and venous tissue segmentation results are merged to obtain a first venous tissue separation result. Then, the arterial tissue segmentation results and venous tissue classification results are merged to obtain a second venous tissue separation result. Then, based on the first and second venous tissue separation results, the venous tissue lumen centerline is extracted and identified using image processing technology to obtain the lumen centerline identification result. Then, based on the lumen centerline identification result, the venous lumen morphology parameters are calculated to obtain the venous stenosis identification result. Thus, for the same image sequence, arterial, venous, thrombus, calcification, and other tissue segmentation results can be obtained simultaneously. Based on multi-sequence CT image arterial and venous segmentation, even in the case of contrast agent interference, effective separation of arteries and veins can be achieved for venous stenosis identification, greatly reducing the diagnosis and treatment time and improving the accuracy of the evaluation results.

[0078] Figure 2 A schematic diagram of arteriovenous segmentation of the lower limbs according to an embodiment of this application is shown.

[0079] like Figure 2 The image shows a schematic diagram of arteriovenous segmentation in the lower extremities. Figure 2 In the diagram, arterial and venous tissues are displayed using different color labels. The various parts of the arterial tissue are displayed together with a single red label, while the various parts of the venous tissue are categorized using other different color labels, with each color representing a different part of the vein.

[0080] For example, after performing vein abnormality identification based on at least one of the arteriovenous tissue segmentation results and the vein abnormality tissue segmentation results, morphological parameters can be calculated based on the vein abnormality identification results. For example, morphological parameters of vein abnormality tissue can be calculated based on the vein thrombosis identification results and the vein calcification identification results. The morphological parameters of vein abnormality tissue include at least one of coordinates, length, volume, mean, and volume percentage.

[0081] Specifically, based on the results of third tissue segmentation or venous thrombosis identification, parameters such as the location, length, volume, mean, variance, and volume percentage of thrombi in veins can be calculated; and based on the results of fourth tissue segmentation or venous calcification identification, parameters such as the location, length, volume, mean, and volume percentage of calcifications can be calculated, thereby obtaining morphological parameters of abnormal venous tissues.

[0082] Among them, the abnormal morphological parameters of vein tissue and the morphological parameters of vein lumen obtained above constitute the morphological parameters of vein, which can be used as the basis for preoperative approach assessment or postoperative efficacy assessment of inferior vena cava occlusion.

[0083] In the technical solution of this application embodiment, the abnormal morphological parameters of venous tissue are calculated based on the venous thrombosis identification results and the venous calcification identification results. The calculation results are then used as the basis for preoperative approach assessment or postoperative efficacy assessment of inferior vena cava occlusion, thereby further improving the accuracy of the assessment results.

[0084] Figure 3 A schematic diagram showing the vein morphology parameter results of an embodiment of this application is illustrated.

[0085] like Figure 3 As shown, the calculation results of vein morphology parameters obtained by this method can be displayed in any interface format, such as charts or text. It supports displaying corresponding parameter information categorized by vein or vein location, and can also be displayed using image markers. (Reference) Figure 3 The chart shows the tissue direction in the first column, the vein and its specific location in the second column, the specific location of each part of the vein in the third column, and the remaining columns are the morphological parameters calculated for each part.

[0086] Therefore, this method achieves precise separation of lower extremity arteries and veins based on multi-sequence CT images (including plain scan, arterial phase, and venous phase), eliminating the influence of contrast agent interference. Based on the identified venous vessel location information, it enables precise detection of thrombi and calcifications on plain scan images, providing qualitative and quantitative assessments of lesions and helping physicians rule out unsuitable puncture approaches. Simultaneously, based on the segmentation and classification results of the inferior vena cava and renal vein, it provides quantitative assessment parameters to assist physicians in selecting the optimal filter size and the safe area for filter fixation, demonstrating innovation and clinical application value.

[0087] For example, the vein anomaly identification results include vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results; based on the vein anomaly identification results, an inferior vena cava occlusion assessment is performed to obtain an assessment result, including at least one of the following: for example, based on at least one of the vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results, an effective path for inferior vena cava occlusion is assessed to obtain a target occlusion path as the assessment result; or, for example, based on the vein thrombosis identification results, a safe area for the release of the inferior vena cava occluder is assessed to obtain a target occlusion area as the assessment result.

[0088] For example, based on the results of vein anomaly identification, the inferior vena cava occlusion assessment is performed to obtain the assessment results. The assessment also includes: evaluating the size of the inferior vena cava occluder based on the morphological parameters of the vein lumen to obtain the target size as the assessment result.

[0089] Specifically, the preoperative assessment for inferior vena cava occlusion based on the results of venous anomaly identification includes the following parts: (1) Confirm an effective surgical approach Based on the results of the third, fourth, and fifth tissue segmentation, as well as the third tissue classification, identify any abnormalities (e.g., stenosis / calcification / thrombosis) along the left / right path from the lower extremity veins to the inferior vena cava. Confirm the presence of abnormalities in each venous segment, and then perform a risk assessment for each vein. The assessment criteria include, but are not limited to, any of the following: If the external iliac vein has stenosis / calcification / thrombosis, the risk level of the lower extremity vein approach on that side is high, and this approach is not recommended for surgery.

[0090] If the external iliac vein is not narrowed / calcified / thrombotic, but the common femoral vein is narrowed / calcified / thrombotic, then the external iliac vein is used as the surgical approach.

[0091] If there is stenosis / calcification / thrombosis in the inferior vena cava, the risk level of the lower extremity vein approach is high, and it is not recommended to choose the lower extremity vein as the surgical approach.

[0092] If the risk level of both lower extremity vein approaches is high, it is not recommended to choose the lower extremity vein as the surgical approach.

[0093] Therefore, the path with the lowest risk level was selected as the best surgical approach, i.e., the target occlusion path.

[0094] (2) Calculate the size of the plug and the safe zone for plug release. Based on the venous lumen morphology parameters obtained above, the lumen morphology information of the inferior vena cava is acquired, thereby selecting the appropriate occluder size, which yields the evaluation result of the target size.

[0095] Based on the principles of the "Expert Consensus on the Standardized Procedures for Inferior Vein Filter Placement and Removal," the safe zone for occluder release can be determined according to the results of venous thrombosis identification. First, based on the segmentation and classification results of the renal vein and inferior vena cava mentioned above, the opening positions of the left / right renal veins and the effective range of the inferior vena cava are calculated. For example, the nearest position connected to the aorta can be determined based on the connectivity of each part, which is the opening position.

[0096] Then, assess whether there is thrombus at the renal vein opening. If so, record the area above the renal vein opening as the safe area for occluder release; or assess whether there is thrombus in the inferior vena cava within a certain range below the renal vein opening. If so, record the area above the renal vein opening as the safe area for occluder release; if neither of the above is present, select the area below the lower edge of the renal vein opening and within the effective range of the inferior vena cava as the safe area for occluder release, thus obtaining the assessment result of the target occlusion area.

[0097] In the technical solution of this application embodiment, the effective path for inferior vena cava occlusion is evaluated based on at least one of the results of venous thrombosis identification, venous calcification identification, and venous stenosis identification, and the target occlusion path is obtained as the evaluation result; the safe area for the release of the inferior vena cava occluder is evaluated based on the venous thrombosis identification result, and the target occlusion area is obtained as the evaluation result; the size of the inferior vena cava occluder is evaluated based on the venous lumen morphology parameters, and the target size is obtained as the evaluation result. This allows for fully automated segmentation and classification of arteries, veins, and abnormal tissues based on multi-sequence CT images, providing doctors with precise quantitative and qualitative analysis of the inferior vena cava, renal veins, and lower limb veins, reducing manual operation. By combining image processing with deep learning technology, automatic optimal recommendation of puncture approaches is achieved, assisting doctors in developing personalized surgical plans and efficacy evaluations, significantly reducing treatment time, lowering surgical risks, and improving patient survival rates and quality of life.

[0098] Figure 4 A schematic diagram showing the results of the inferior vena cava occlusion risk assessment according to an embodiment of this application is illustrated.

[0099] like Figure 4 As shown, this method supports displaying risk assessment results in any chart format, or via image annotations. (Reference) Figure 4 The table has the following columns: the first column shows the type and risk level of lesions in the venous tissue; the first row shows the direction of the venous tissue; the second row shows the specific location of the venous tissue; the third to fifth rows show the identification results of each lesion; and the sixth row shows the risk assessment result (high or low).

[0100] For example, a structured assessment report can also be generated based on the results of various segmentation, classification, quantitative and qualitative analyses. For instance, a structured report can be generated based on at least two of the following: arteriovenous tissue segmentation results, abnormal venous tissue segmentation results, abnormal venous identification results, venous tissue classification results, assessment results, venous lumen morphological parameters, and abnormal venous tissue morphological parameters.

[0101] Specifically, all the segmentation results, classification results, quantitative analysis results, and qualitative analysis results obtained above can be displayed in any combination. Key information can be output in the form of automated structured reports, and key images can be output in the form of automatic batch processing. Thus, by combining image processing with deep learning technology, automatic optimal recommendation of puncture approach and quantitative analysis of inferior vena cava and renal vein can be achieved, and structured evaluation reports can be automatically generated.

[0102] In the technical solution of this application embodiment, a structured report is generated based on at least two of the following: arteriovenous tissue segmentation results, abnormal venous tissue segmentation results, abnormal venous identification results, venous tissue classification results, evaluation results, venous lumen morphological parameters, and abnormal venous tissue morphological parameters. Automated full-process evaluation can greatly reduce treatment time and reduce labor input costs. Accurate full-process evaluation can greatly reduce surgical risks and improve patient survival rate and quality of life.

[0103] Figure 5 A block diagram of a medical evaluation device according to an embodiment of this application is shown.

[0104] like Figure 5 As shown, this application provides a medical assessment device 500, which includes: The acquisition module 510 is used to acquire target medical images with multiple sequences.

[0105] The segmentation module 520 is used to segment a multi-sequence target medical image and determine at least one of the segmentation results of arteriovenous tissue and abnormal venous tissue.

[0106] The identification module 530 is used to identify vein abnormalities based on at least one of the arteriovenous tissue segmentation results and the vein abnormality tissue segmentation results, and obtain vein abnormality identification results.

[0107] The assessment module 540 is used to assess inferior vena cava occlusion based on the results of venous anomaly identification and obtain assessment results.

[0108] For example, the vein abnormality identification results include vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results; the evaluation module 540 is further configured to: evaluate the effective path for inferior vena cava occlusion based on at least one of the vein thrombosis identification results, vein calcification identification results, and vein stenosis identification results, and obtain the target occlusion path as the evaluation result; and evaluate the safe area for release of the inferior vena cava occluder based on the vein thrombosis identification results, and obtain the target occlusion area as the evaluation result.

[0109] For example, the arteriovenous tissue segmentation results include arterial tissue segmentation results and venous tissue segmentation results, and the venous abnormal tissue segmentation results include thrombus tissue segmentation results and calcified tissue segmentation results; the identification module 530 is further configured to: perform venous thrombosis identification based on the venous tissue segmentation results and thrombus tissue segmentation results to obtain venous thrombosis identification results; perform venous calcification identification based on the venous tissue segmentation results and calcified tissue segmentation results to obtain venous calcification identification results; and perform venous stenosis identification based on the arterial tissue segmentation results and venous tissue segmentation results to obtain venous stenosis identification results.

[0110] For example, the target medical image of the multi-sequence includes plain scan image, arterial phase image, and venous phase image; based on the venous tissue segmentation result and the thrombus tissue segmentation result, venous thrombosis identification is performed to obtain the venous thrombosis identification result, including: classifying the venous tissue of the target medical image of the multi-sequence according to the venous tissue segmentation result and the venous phase image to obtain the venous tissue classification result; and identifying the venous thrombosis site of the target medical image of the multi-sequence according to the venous tissue classification result and the thrombus tissue segmentation result to obtain the venous thrombosis identification result.

[0111] For example, based on the vein tissue segmentation results and calcified tissue segmentation results, vein calcification is identified to obtain vein calcification identification results, including: based on the vein tissue classification results and calcified tissue segmentation results, vein calcification sites are identified in the target medical image of multiple sequences to obtain vein calcification identification results.

[0112] For example, based on the arterial tissue segmentation results and the venous tissue segmentation results, venous stenosis identification is performed to obtain venous stenosis identification results, including: merging the arterial tissue segmentation results and the venous tissue segmentation results according to the positional relationship between the arterial tissue and the venous tissue in the multi-sequence target medical image to obtain a first venous tissue separation result, and merging the arterial tissue segmentation results and the venous tissue classification results to obtain a second venous tissue separation result; extracting and identifying the lumen centerline of the venous tissue using image processing technology based on the first venous tissue separation result and the second venous tissue separation result to obtain a lumen centerline identification result; and calculating the venous lumen morphological parameters based on the lumen centerline identification result to obtain the venous stenosis identification result, wherein the venous lumen morphological parameters include at least one of diameter, area, mean, variance, and stenosis rate.

[0113] For example, the segmentation module 520 is further configured to: perform arterial tissue segmentation on the multi-sequence target medical image based on the arterial phase image to obtain arterial tissue segmentation results; and perform venous tissue segmentation on the multi-sequence target medical image based on the venous phase image to obtain venous tissue segmentation results; and perform venous thrombosis segmentation and venous calcification segmentation on the multi-sequence target medical image based on the plain scan image and the venous tissue segmentation results to obtain thrombosis tissue segmentation results and calcification tissue segmentation results.

[0114] For example, after performing vein abnormality identification based on at least one of the arteriovenous tissue segmentation results and the vein abnormality tissue segmentation results, the device 500 further includes a morphological parameter calculation module for: calculating the morphological parameters of the vein abnormality tissue based on the vein thrombosis identification results and the vein calcification identification results, wherein the morphological parameters of the vein abnormality tissue include at least one of coordinates, length, volume, mean, and volume percentage.

[0115] For example, the evaluation module 540 is further configured to: evaluate the size of the inferior vena cava occluder based on the morphological parameters of the vein lumen, and obtain the target size as the evaluation result.

[0116] For example, the device 500 further includes a generation module for generating a structured report based on at least two of the following: arteriovenous tissue segmentation results, abnormal venous tissue segmentation results, abnormal venous identification results, venous tissue classification results, evaluation results, venous lumen morphological parameters, and abnormal venous tissue morphological parameters.

[0117] Figure 6 A schematic diagram of an electronic device according to an embodiment of this application is shown.

[0118] This application provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the method in any of the above embodiments.

[0119] like Figure 6 As shown, for ease of understanding, an embodiment of this application illustrates a specific electronic device 600.

[0120] Electronic device 600 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 600 may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0121] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 608.

[0122] Multiple components in electronic device 600 are connected to I / O interface 605. These components include: input unit 606, such as a keyboard or mouse; output unit 607, such as various types of displays or speakers; storage unit 608, such as a disk or optical disk; and communication unit 609, such as a network interface card (NIC), modem, or wireless transceiver. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0123] The computing unit 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods described above. For example, in some embodiments, any one or more of the methods described above can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of any one or more of the methods described above can be performed. Alternatively, in other embodiments, the computing unit 601 can be configured to perform any one or more of the methods described above by any other suitable means (e.g., by means of firmware).

[0124] This application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method in any of the above embodiments.

[0125] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this application, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0126] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any way. This application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A medical assessment method, characterized in that, The method includes: Acquire target medical images with multiple sequences; The target medical image of the multi-sequence is segmented to determine at least one of the segmentation results of arteriovenous tissue and segmentation results of abnormal venous tissue. Based on at least one of the arteriovenous tissue segmentation results and the abnormal vein tissue segmentation results, vein abnormality identification is performed to obtain vein abnormality identification results. Based on the results of the vein abnormality identification, an assessment of inferior vena cava occlusion was performed, and the assessment results were obtained.

2. The medical assessment method according to claim 1, characterized in that, The venous anomaly identification results include venous thrombosis identification results, venous calcification identification results, and venous stenosis identification results; based on the venous anomaly identification results, the inferior vena cava occlusion assessment is performed to obtain the assessment results, which include at least one of the following: Based on at least one of the venous thrombosis identification results, the venous calcification identification results, and the venous stenosis identification results, the effective path for inferior vena cava occlusion is evaluated, and the target occlusion path is obtained as the evaluation result. Based on the venous thrombosis identification results, the safe area for the release of the inferior vena cava occluder is evaluated, and the target occlusion area is obtained as the evaluation result.

3. The medical assessment method according to claim 2, characterized in that, The arteriovenous tissue segmentation results include arterial tissue segmentation results and venous tissue segmentation results; the abnormal venous tissue segmentation results include thrombus tissue segmentation results and calcified tissue segmentation results; the venous abnormality identification is performed based on at least one of the arteriovenous tissue segmentation results and the abnormal venous tissue segmentation results to obtain venous abnormality identification results, including at least one of the following: Based on the vein tissue segmentation results and the thrombus tissue segmentation results, venous thrombosis is identified to obtain the venous thrombosis identification results; Based on the vein tissue segmentation results and the calcified tissue segmentation results, vein calcification is identified to obtain the vein calcification identification results; Based on the arterial tissue segmentation results and the venous tissue segmentation results, venous stenosis is identified to obtain the venous stenosis identification results.

4. The medical assessment method according to claim 3, characterized in that, The multi-sequence target medical images include plain scan images, arterial phase images, and venous phase images; the venous thrombosis identification is performed based on the venous tissue segmentation results and the thrombus tissue segmentation results to obtain the venous thrombosis identification results, including: Based on the vein tissue segmentation results and the venous phase images, the target medical images of the multi-sequences are classified into vein tissues to obtain vein tissue classification results. Based on the vein tissue classification results and the thrombus tissue segmentation results, the target medical image of the multi-sequence is used to identify the venous thrombosis site, and the venous thrombosis identification result is obtained.

5. The medical assessment method according to claim 3, characterized in that, The process of identifying vein calcifications based on the vein tissue segmentation results and the calcified tissue segmentation results, to obtain the vein calcification identification results, includes: Based on the vein tissue classification results and the calcified tissue segmentation results, the vein calcification sites are identified in the multi-sequence target medical image to obtain the vein calcification identification results.

6. The medical assessment method according to claim 3, characterized in that, The process of identifying venous stenosis based on the arterial tissue segmentation results and the venous tissue segmentation results, to obtain the venous stenosis identification results, includes: Based on the positional relationship between arterial and venous tissues in the target medical images of the multi-sequence, the segmentation results of the arterial tissue and the segmentation results of the venous tissue are merged to obtain a first venous tissue separation result, and the segmentation results of the arterial tissue and the classification results of the venous tissue are merged to obtain a second venous tissue separation result; Based on the first vein tissue separation result and the second vein tissue separation result, the lumen centerline of the vein tissue is extracted and identified by image processing technology to obtain the lumen centerline identification result. Based on the identification result of the lumen centerline, the morphological parameters of the vein lumen are calculated to obtain the identification result of the vein stenosis. The morphological parameters of the vein lumen include at least one of diameter, area, mean, variance, and stenosis rate.

7. The medical assessment method according to claim 4, characterized in that, The step of segmenting the target medical image of the multi-sequence array to determine at least one of the arteriovenous tissue segmentation result and the abnormal venous tissue segmentation result includes: Based on the arterial phase image, the target medical image of the multi-sequence is segmented into arterial tissue to obtain the arterial tissue segmentation result; and based on the venous phase image, the target medical image of the multi-sequence is segmented into venous tissue to obtain the venous tissue segmentation result. Based on the plain scan image and the vein tissue segmentation results, the target medical image of the multi-sequence is segmented into vein thrombosis and vein calcification respectively to obtain the thrombosis tissue segmentation results and the calcification tissue segmentation results.

8. The medical assessment method according to any one of claims 1-7, characterized in that, After performing vein abnormality identification based on at least one of the arteriovenous tissue segmentation results and the abnormal vein tissue segmentation results to obtain the vein abnormality identification result, the method further includes: Based on the venous thrombosis identification results and the venous calcification identification results, venous abnormal tissue morphological parameters are calculated, wherein the venous abnormal tissue morphological parameters include at least one of coordinates, length, volume, mean, and volume percentage.

9. The medical assessment method according to claim 6, characterized in that, The process of assessing inferior vena cava occlusion based on the venous anomaly identification results, and obtaining the assessment results, also includes: The size of the inferior vena cava occluder is evaluated based on the venous lumen morphology parameters, and the target size is obtained as the evaluation result.

10. The medical assessment method according to claim 9, characterized in that, The method also includes: A structured report is generated based on at least two of the following: the arteriovenous tissue segmentation result, the abnormal venous tissue segmentation result, the abnormal venous identification result, the venous tissue classification result, the evaluation result, the venous lumen morphology parameters, and the abnormal venous tissue morphology parameters.

11. A medical assessment device, characterized in that, The device includes: The acquisition module is used to acquire target medical images with multiple sequences; The segmentation module is used to perform target segmentation on the multi-sequence target medical image and determine at least one of the segmentation results of arteriovenous tissue and segmentation results of abnormal venous tissue. The identification module is used to identify vein abnormalities based on at least one of the arteriovenous tissue segmentation results and the abnormal vein tissue segmentation results, and to obtain a vein abnormality identification result. The evaluation module is used to evaluate the inferior vena cava occlusion based on the venous anomaly identification results and obtain the evaluation results.

12. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-10.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-10.

Images