CN-119068137-B - Liver 3D printing model construction method, system and equipment based on image data

CN119068137BCN 119068137 BCN119068137 BCN 119068137BCN-119068137-B

Abstract

The application provides a liver 3D printing model construction method, a liver 3D printing model construction system and liver 3D printing model construction equipment based on image data, which comprise the steps of obtaining CT image data and MRI image data of an individual patient; the method comprises the steps of carrying out multi-mode image segmentation on CT image data and MRI image data to obtain structure segmentation images, fusing the structure segmentation images to obtain liver and tumor anatomical structure models, carrying out three-dimensional reconstruction on the liver and tumor anatomical structure models to obtain target three-dimensional visual models, and carrying out 3D printing on the target three-dimensional visual models. The application combines CT and MRI image modes, and can obtain more comprehensive and accurate liver anatomical information. CT provides clear liver parenchyma and vascular images, MRI provides more sensitive soft tissue contrast and tumor detection. The method and the device can make up for the defects of a single image technology and improve the fidelity and information quantity of the model.

Inventors

WU LIMING
WANG YI
LIN BINGYI
HONG DONGSHENG

Assignees

浙江大学

Dates

Publication Date: 20260505
Application Date: 20240705

Claims (10)

1. The method for constructing the liver 3D printing model based on the image data is characterized by comprising the following steps of: acquiring CT image data and MRI image data of an individual patient; Performing multi-mode image segmentation on the CT image data and the MRI image data to obtain a structure segmentation image; reconstructing a vascular network based on arterial phase and venous phase data in the CT image data; Fusing the structure segmentation images to obtain anatomical structure models of livers and tumors; Registering the reconstructed vascular network with the liver and tumor anatomical structure model to obtain an integral liver model containing a fine vascular network; Performing three-dimensional reconstruction on the integral liver model to obtain a target three-dimensional visual model; 3D printing is conducted on the target three-dimensional visual model; the reconstruction of the vascular network comprises the following steps: Performing blood vessel segmentation on the CT image of the arteriovenous phase, extracting hepatic artery by taking 250-450HU as a threshold value, and extracting hepatic vein by 100-250 HU; morphological treatment is carried out on the segmentation result, and 1 multiplied by 1 structural elements are adopted to expand the blood vessel so as to fill the fracture of the lumen; Fusing portal vein and hepatic vein, smoothing surface, and generating a complete hepatic vessel tree; referring to the Couinaud segmentation system, a 3D printed liver model is divided into I-VIII regions, each region is marked by labels with different colors, the three-dimensional model is further subdivided into subsections on the basis of eight sections of the Couinaud, and the VIII section is divided into two subsections, namely VIIIa and VIIIb, which respectively represent the upper branch drainage region and the lower branch drainage region of the VIII section vein.
2. The method for constructing a liver 3D printing model based on image data according to claim 1, wherein the multi-modal image segmentation is performed on the CT image data and the MRI image data to obtain segmented images, comprising: And carrying out threshold segmentation processing on the CT image data according to a plurality of preset threshold ranges to obtain a threshold structure segmentation image, and extracting an interested region of the MRI image data according to a region growing method to obtain an interest structure segmentation image, wherein the structure segmentation image comprises the interest structure segmentation image and the threshold structure segmentation image.
3. The method of claim 2, wherein the structure segmentation image comprises a segmentation image of a tumor structure, and the segmentation of the tumor structure segmentation image comprises: Extracting a high-density region from the CT image by threshold segmentation to serve as a tumor candidate region; Highlighting the difference between a tumor region and normal liver tissue in the T1 weighted and T2 weighted images of the MRI by using a dynamic contrast enhancement and diffusion weighted imaging technique; and (3) carrying out spatial registration on tumor candidate areas extracted from the CT and MRI images, extracting accurate boundaries of tumors through morphological operation and connectivity analysis, and segmenting to obtain segmented images of tumor structures.
4. A method of constructing a 3D printing model of a liver based on image data as claimed in claim 3, wherein the structure segmentation image comprises a segmentation image of a liver tissue structure, and the method comprises the steps of: And fusing the segmented images of the liver tissue structure and the tumor structure by using a deep learning algorithm to generate a complete liver and tumor anatomical structure model, wherein the deep learning algorithm takes U-Net as a basic framework.
5. The method for constructing a liver 3D printing model based on image data according to claim 1, wherein the 3D printing comprises: And selecting proper printing materials and printing process parameters, rapidly manufacturing a target three-dimensional visual model corresponding to the liver and the tumor on a 3D printer, and cleaning and curing the printed model to obtain a final 3D printing model of the liver and the liver tumor.
6. The method for constructing a 3D printing model of a liver based on image data according to claim 1, wherein three-dimensionally reconstructing the anatomical model of the liver and the tumor to obtain a three-dimensional visualization model of the target comprises: an organ structure of the liver and tumor anatomical structure model is segmented by an image segmentation algorithm based on deep learning; Extracting three-dimensional organ features of an organ structure based on a deep learning image feature extraction algorithm; And generating a target three-dimensional visual model according to the extracted three-dimensional features of the organ and a three-dimensional reconstruction algorithm.
7. The method for constructing a 3D printing model of a liver based on image data according to claim 1, wherein three-dimensionally reconstructing the anatomical model of the liver and the tumor to obtain a three-dimensional visualization model of the target comprises: Fusing the characteristics of CT and MRI two-mode data in the encoder part, and reconstructing a continuous smooth three-dimensional structure of liver and tumor in the decoder part; Extracting geometric features of the liver surface and the tumor surface and topological features of pipeline structures such as blood vessels, bile ducts and the like in the liver according to the fused anatomical structure model of the liver and the tumor; assembling the reconstructed liver, tumor surface model and pipeline structure model to obtain a high-fidelity three-dimensional visualized model of the liver and the tumor; And converting the personalized three-dimensional visualization model of the liver and the tumor into three-dimensional printing model data.
8. The method for constructing the liver 3D printing model based on the image data, which is disclosed in claim 7, is characterized in that in extracting geometric features of the liver surface and the tumor surface and topological features of pipeline structures such as blood vessels and bile ducts in the liver, the reconstruction of grids on the liver and the tumor surface and the skeleton extraction of the pipeline structures are respectively realized by adopting Marching Cubes algorithm and Dijkstra shortest path algorithm.
9. A liver 3D printing model construction system based on image data of the method of any one of claims 1 to 8, comprising: The image data acquisition module is used for acquiring CT image data and MRI image data of the patient individual; The image segmentation module is used for carrying out multi-mode image segmentation on the CT image data and the MRI image data to obtain a structure segmentation image; The image fusion module is used for fusing the structure segmentation images to obtain a liver and tumor anatomical structure model; the three-dimensional reconstruction module is used for carrying out three-dimensional reconstruction on the anatomical structure model of the liver and the tumor to obtain a target three-dimensional visual model; and the printing module is used for 3D printing of the target three-dimensional visual model.
10. An electronic device comprising a processor, a memory for storing instructions, and a transceiver for communicating with other devices, the processor for executing the instructions stored in the memory to cause the electronic device to perform the method of any one of claims 1-8.

Description

Liver 3D printing model construction method, system and equipment based on image data Technical Field The invention relates to the field of medical images, in particular to a liver 3D printing model construction method, system and equipment based on image data. Background The medical image data are converted into the 3D printing model, so that a doctor can be assisted in accurate preoperative planning and intra-operative navigation, the medical image data can be used for medical teaching, and the understanding of medical students and inpatients on liver anatomical structures and diseases can be improved. There have been studies to make a normal liver model for medical teaching and training using 3D printing techniques. The models can show three-dimensional structural characteristics of anatomical morphology, couinaud partition, vascular distribution running and the like of the liver, and can enable learners to comprehensively and multi-level observe and know the internal structure of the liver by carrying out operations such as assembly, disassembly, fault incision and the like on the models, so that the limitations of traditional autopsy and two-dimensional map teaching are overcome to a certain extent. However, most of the existing 3D-printed liver models are created based on image data of normal persons, and cannot simulate liver tumor. The liver tumor has obvious individual differences in size, position, morphology, blood supply and the like, and has different relations with surrounding normal liver tissues and blood vessels. Lack of tumor reproduction in the 3D printing model makes it difficult to meet clinical diagnosis and teaching requirements. On the other hand, diagnosis and operation planning of liver tumor in clinic at present mainly depend on imaging examination such as CT, MRI and the like. However, these image data are two-dimensional plane images, which are difficult to intuitively and stereoscopically show the spatial structure relationship of liver and tumor, and also difficult to satisfy the clinical diagnosis and teaching requirements. Disclosure of Invention The application provides a liver 3D printing model construction method, a liver 3D printing model construction system and liver 3D printing model construction equipment based on image data, which can intuitively and three-dimensionally display the spatial structure relation of livers and tumors so as to meet the requirements of clinical diagnosis and treatment and teaching. The first aspect of the application provides a liver 3D printing model construction method based on image data; A liver 3D printing model construction method based on image data comprises the following steps: acquiring CT image data and MRI image data of an individual patient; carrying out multi-mode image segmentation on the CT image data and the MRI image data to obtain a structure segmentation image; Fusing the structure segmentation images to obtain anatomical structure models of livers and tumors; performing three-dimensional reconstruction on the liver and tumor anatomical structure model to obtain a target three-dimensional visual model; and 3D printing is carried out on the target three-dimensional visual model. Preferably, the multi-mode image segmentation is performed on the CT image data and the MRI image data to obtain segmented images, including: And carrying out threshold segmentation processing on the CT image data according to a plurality of preset threshold ranges to obtain a threshold structure segmentation image, and extracting an interested region of the MRI image data according to a region growing method to obtain an interest structure segmentation image, wherein the structure segmentation image comprises the interest structure segmentation image and the threshold structure segmentation image. Preferably, before the multi-modal image segmentation is performed on the CT image data and the MRI image data to obtain the structure segmentation image, the method includes: the resolution, pixel size and the coordinate system of the CT image data and the MRI image data are adjusted to be consistent. Preferably, the fusion of the structure segmentation images to obtain the anatomical model of liver and tumor comprises: and inputting the structure segmentation image into a deep learning image fusion model to obtain a liver and tumor anatomical structure model. Preferably, the image fusion model for deep learning is a convolutional neural network model or a generated countermeasure network model. Preferably, the three-dimensional reconstruction of the anatomical structure model of the liver and the tumor to obtain the three-dimensional visualization model of the target comprises the following steps: an organ structure of the liver and tumor anatomical structure model is segmented by an image segmentation algorithm based on deep learning; Extracting three-dimensional organ features of an organ structure based on a deep learning image feature extraction a