| Three-dimensional digital modeling is a key step in bionic manufacturing of human tissues and organs, which has been a hot topic in the field of computer aided bionic manufacturing. It can help doctors to improve the scientificity and accuracy of medical diagnosis, and has been widely used in plastic and artificial limb, surgical simulation, medical anatomy teaching and other fields. With the development of medical technology and the improvement of people’s living standard, minimally invasive concept gradually win support among the people. Correspondingly, the content of3D digital modeling has also been further refined, and gradually develop from macroscopic to microscopic direction, the overall level to the local level. The3D digital modeling is not limited to the individual and system level, but further develop to the organ level, namely single organ three-dimensional digital modeling. It can give us a comprehensive understanding of single organ blood supply and distribution system, and can help us deeply and vividly understand the occurrence and development of disease. At present, many human single organ model has been constructed both at home and abroad, which mainly focus on the surface reconstruction of tissues and organs, and neglect the extraction and expression of its the internal structure and information, such as the brain, heart, lung, liver, kidney, bone and joint, eye, but there is no3D digital modeling research reports of internal reproductive organs. so it is necessary to study the3D digital modeling of internal reproductive organs. However, high quality dataset collection is the basis of3D digital modeling. Therefore, studying the method for obtaining single organ thin-layer dataset, and constructing the three-dimensional digital model of internal reproductive organs is very necessary.The research is totally divided into five parts. The first part:collect MRI scanning dataset, CT scanning dataset, and milling slice dataset by using MRI, CT and digital technology. The second part:construct three-dimensional digital model of normal human uterine, ovarian, fallopian tube, part of the vagina, uterine artery vascular network, uterine vein vascular network based on MRI thin-layer scanning dataset. The third part:construct three-dimensional digital model of normal human uterine, ovarian, fallopian tube, part of the vagina, uterine artery vascular network, uterine vein vascular network based on CT thin-layer scanning dataset. The fourth part:Firstly, improve and optimize the image registration method based on Photoshop and Matlab software, and complete the automatic registration of milling slice slice dataset; And then construct the3D digital models of human normal in-vitro uterus, tubal, ovarian, part of the vagina, uterine artery vascular network, uterine vein vascular network, the endometrium and myometrium based on the milling slice dataset of human normal in-vitro internal reproductive organs after registration, by using Photoshop, ACDSee and Mimics; The fifth part:compare the precision and breadth of3D digital models constructed by MRI scanning dataset, CT scanning dataset and milling slice dataset, and confirm the method suitable for thin-layer dataset collection of the internal reproductive organs, so as to provide high quality anatomical materials for clinical, research and teaching, and provide new ideas and methods for thin-layer dataset collection of other single organs.[Purpose]Acquire MRI thin-layer scannig dataset, CT thin-layer scanning dataset and milling slice dataset of human normal in-vitro internal reproductive organ, so as to provide the anatomical basis for the further study.[Method]Acquire2human normal adult female cadaver specimens, and integrallty cut off2specimens, Including the uterus, tubal, ovarian and part of the vagina; mix30%gelatin, amount of10%soluble starch and lead oxide and amount of blue-black ink and stir evenly, so as to generate gelatin infusion; select one specimen, separat the trunk of bilateral uterine artery, insert the glass tube of appropriate diameter and ligature the proximal end of glass tube. Separat the trunk of bilateral uterine vein on the other uterus, insert the glass tube of appropriate diameter and ligature the proximal end of glass tube; scan the uterus by using64slice spiral CT and Philips3.0T MRI before perfusion, so as to understand the situation of the uterus; perfuse amount of gelatin solution into uterine artery vascular network and uterine vein vascular network through the tube and control perfusion pressure with certain pressure measuring device; again CT and MRI scanning after reperfusion, and obtain CT and MRI scanning dataset of normal in-vitro reproductive organs; put the sample into-70℃refrigerator for72h, and freeze the uterus; take an appropriate amount of gelatin in30%, the amount of blue-black ink, mixing evenly, and generate into gelatin embedding agent; put the frozen specimens into special containers, and embed the uterus by Gypsum; acquisition set parameters of cutting machine; start up the machine, equidistant cut after setting the feed volume to0.2mm, and obtain the sectional images by using professional camera, so as to construct the milling slice dataset of internal reproductive organst.[Results]1.based on the MRI, CT and digital technology, we successfully obtained MRI scanning dataset, CT scanning dataset and milling slice dataset of the human normal internal reproductive organs.2.MRI, CT thin-layer scanning dataset after perfusionThe amout of MRI thin-layer scanning image of uterus perfused through uterine artery is20, and the amout of MRI thin-layer scanning image of uterus perfused through uterine vein is20; The amout of CT thin-layer scanning image of uterus perfused through uterine artery is20, and the amout of CT thin-layer scanning image of uterus perfused through uterine vein is20.3.The miling dataset of human normal in-vitro internal reproductive organs is composed of510consecutive sectional image, and the sectional interval is0.2mm.(1)original dataset:in order to maximized retain the original information of dataset, we use the uncompressed RAW format. The resolution is5184x3456, and the total amount of dataset is16.6G.(2)decoded dataset:in order to make it suitable for browsing software in most image processing operation, we decode it into510TIF format dataset, with a resolution of5184x3456pixels, and the total amount of dataset is51.0G.(3)compressed dataset:in order to make it suitable for PC operation, we compress the TIF dataset into JPEG format dataset, with a resolution of5184x3456, and the total amount of dataset is7.4G[Conclusion]Using MRI, CT and digital technology, we can successfully obtain MRI scanning dataset, CT scanning dataset and milling slice dataset of the human normal internal reproductive organs.[Purpose]3D reconstruct the human normal in-vitro uterine MRI scanning dataset.[Method]Use RadiAnt DICOM Viewer to access the dataset; import dataset into Mimics, select the appropriate threshold after positioning and interpolation processing, construct the3D digital model of human in-vitro internal reproductive organs, uterine artery vascular network and uterine vein vascular network by regional reinforcement, and observe the characteristics of the model; save the results for JPEG format.[Results]1.based on MRI thin-layer scanning dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina by using Mimics, but this model is relatively rough and with poor precisionr, low resolution.2. based on MRI thin-layer scanning dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of human normal in-vitro "uterine artery vascular network" and "uterine vein vascular network", but actually soft tissue.[Conclusion]We can construct a relatively rough3D digital model of the reproductive organs, but can not construct a three-dimensional digital model of uterine artery and uterine vein vascular network. [Purpose]3D reconstruct the human normal in-vitro uterine CT scanning dataset.[Method]Use RadiAnt DICOM Viewer to access the dataset; import dataset into Mimics, select the appropriate threshold after positioning and interpolation processing, construct the3D digital model of human in-vitro internal reproductive organs, uterine artery vascular network and uterine vein vascular network by regional reinforcement, and observe the characteristics of the model; save the results for JPEG format.[Results]1.based on CT thin-layer scanning dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina by using Mimics. The model is relatively smooth and with a high precision and resolution.2. based on CT thin-layer scanning dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterine artery vascular network by using Mimics. The model has a high precision and can clearly display the ascending branches of uterine artery, the ovarian branch of uterine artery, uterine fundus branch of uterine artery, ovarian artery and tubal branch of ovarian artery.3.based on CT thin-layer scanning dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterine vein vacular network by using Mimics. The model has a high precision and can clearly display the uterine vein, ovarian branch of uterine vein, uterine fundus branch of ovarian vein and ovarian vein.[Conclusion] We can not only construct a high precision and high resolution3D digital model of the reproductive organs based on CT thin-layer scanning dataset, but can also construct a high precision and high resolution3D digital model of uterine artery vacular network and uterine vein vascular network based on CT thin-layer scanning dataset, which can be used for the three-dimensional digital modeling study of human normal in-vitro internal reproductive organs.[Purpose]improve and optimize the image registration method based on Photoshop and Matlab, and realize the automatic registration of milling slice dataset for the normal human in-vitro internal reproductive organs.[Method]get the positioning rod image and specimen image (uterus, double accessories and part of the vagina) by processing raw images in Photoshop; calculate the centroid coordinate value of four positioning rod in the positioning rod image in the Matlab, and set the mean value of all the positioning rod coordinate values in positioning rod image as a reference coordinate; obtain the projection transformation parameters of each positioning rod image by comparing the positioning rod coordinate values of every positioning rod image and transform the positioning rod image and a positioning rod specimen image according to the parameter; import the corrected image of positioning rod again into Photoshop, and obtain the image only containning the second positioning rod. calculate the coordinate value in Matlab, and cut the specimen positioning rod image after image registration into the same size based on the coordinate values; choose two adjacent images after registration with a large difference in size, overlap the adjacent images before and after the registration, regulate the transparency of one image, and save the synthetic image.[Result]The specimen size difference of two adjacent image before registration is obvious, and the gap after registration is very small.[Conclusion] Image registration method based on Photoshop and Matlab has a high precision, small calculation amount, and is easy to realize. After improvement and optimization, it can be used for registration of human normal in-vitro uterine thin-layer data set.[Purpose]3D reconstruct milling slice dataset of the normal human in-vitro internal reproductive organs.[Method]Import the dataset after registration into Photoshop, switch its background into white; use the magic wand tool to extract uterine artery vascular network and uterine vein vascular network; import the segmented images into Mimics, and obtaine the3D digital model of human in-vitro uterus, fallopian tube, ovary, endometrium, uterine artery vascular network and uterine vein vascular network.[Result]1.Based on milling slice dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina. The model is relatively smooth and has a high precision and resolution.2.Based on milling slice of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterine artery vascular network. The model has a high precision and can clearly display the ascending branches of uterine artery, the ovarian branch of uterine artery, uterine fundus branch of uterine artery, ovarian artery and tubal branch of ovarian artery.3.Based on milling slice dataset of human normal in-vitro internal reproductive organs, we successfully constructed3D digital model of a normal human in-vitro uterine vein vacular network. The model has a high precision and can clearly display the uterine vein, ovarian branch of uterine vein, and uterine fundus branch of ovarian vein.4. Based on milling slice dataset of human normal in-vitro internal reproductive organs, we successfully constructe the3D digital model of human in-vitro endometrial and myometrial.[Conclusion]We can not only construct a high precision and high resolution3D digital model of the reproductive organs, uterine artery vascular network and uterine vein vascular network based on milling slice dataset, but can also construct a high precision and high resolution3D digital model of human in-vitro endometrial and myometrial, which can be used for the three-dimensional digital modeling study of human normal in-vitro internal reproductive organs.[Purpose] compare breadth and accuracy for three-dimensional digital modelsbased on CT scanning dataset, MRI scanning dataset and milling slice dataset and use it to guide the choice of methods of dataset collection.[Method] compare breadth and accuracy for three-dimensional digital modelsbased on CT scanning dataset, MRI scanning dataset and milling slice dataset. [Result]1.We can only construct a3D digital model of normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina, but can not construct a three-dimensional digital model of uterine artery and uterine vein vascular network based on MRI thin-layer scanning dataset;2.We can construct a3D digital model of normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina, uterine artery and uterine vein vascular network based on CT thin-layer scanning dataset;3. We can not only construct a3D digital model of normal human in-vitro uterus, ovarian, fallopian tube, part of the vagina, uterine artery and uterine vein vascular network based on CT thin-layer scanning dataset, but can also the3D digital model of human in-vitro endometrial and myometrial;4.The integrity of three-demensional digital model for the internal reproductive organs based on the milling slice dataset is strongest, and the three-demensional digital model of uterine artery and uterine vein vascular network based on the CT thin-layer scanning dataset is more accuracy.[Conclusion]The three-dimensional digital model of human in-vitro internal reproductive organs based on digital technology is more sophisticated, comprehensive, and can be used for thin-layer dataset collection of human in-vitro internal reproductive organs. |
PDF Full Text Request