Female Pelvic Floor MRI Edge Detection And Digital Model Construction

In the two-dimensional MRI images of the pelvic floor,the positional relationship between various tissues and organs is relatively vague,and it is difficult to obtain a clear sense of three-dimensional space.Therefore,it is of great significance to convert the two-dimensional images into three-dimensional models.The software is generated by manual layer-by-layer filling.The quality of the model constructed by this method is more subject to subjective influence and the overall efficiency is low.Aiming at the construction of female pelvic floor tissue model,combined with relevant optimization algorithms,edge detection,augmented reality and other technologies,a method for edge detection and solid model construction of female pelvic floor MRI is proposed.The specific research contents and conclusions are as follows:(1)Design a parameter optimization model and edge detection method for female pelvic floor nuclear magnetic resonance(NMR)image edge detection results.The optimization model mainly solves the problem of optimizing image-related factors and final evaluation indicator parameters.It uses evaluation indicators to predict and optimize the optimal parameter combination among the related influencing factors.A Canny-SIFT-based edge extraction algorithm is designed to achieve better detection results using the obtained optimal parameter combination.(2)Build a 3D model of the pelvic floor tissue and perform internal and external repairs.Through the study of Dicom files and reverse engineering technology,a method is established to convert edge detection generated 2D images into 3D solid models.Firstly,the Dicom files are parsed to obtain the relevant data required for subsequent 3D model construction,and then rough 3D models are generated with the help of relevant software.Subsequently,the generated models are subjected to internal and external repairs,refined surface and other operations to produce a more realistic3 D model that fits the human body better.(3)Build a multi-method similarity comparison method and set up an AR display platform to complete the virtual scene design.Taking the construction of a 3D model of female pelvic floor tissue and organs as an example,the edge detection effect and similarity of the constructed 3D model are compared using laboratory equipment,in order to verify the rationality of the model generation method.Specifically,this includes qualitative and quantitative analysis of the edge detection effect of multiple algorithms,comparison of the similarity of the model using multiple methods,and finally,using augmented reality technology to fully display the generated tissue model from multiple angles,and package and publish the model on the mobile end for more convenient observation.Through the verification of multi-objective optimized edge detection and solid model construction of female pelvic floor MRI,the results show that the model generated by this method is compared with the standard model.The model has a high degree of similarity and takes a short time to solve the problem of the pelvic floor.The problems of poor MRI edge detection and difficult 3D model generation provide a new method for 3D model generation of female pelvic floor.