Research On MRI Image Segmentation System And Its Realization Based On ITK

As the top of the medical imaging technology,MRI has got rapid development in the last decades years,and been widely used in auxiliary diagnosis of the brain,cardiovascular system and the body's major organs.By use of computer and medical image segmentation technique to extract implicit information from the human body MRI medical imaging,can provide objective data for diagnosis support human disease.Medical image segmentation is the most basic and also one of the most difficult problem in the field of computer vision,it has always been a hot spot in the field of image processing.This paper has studied the MRI brain structure precise segmentation which based on ITK platform and image segmentation system.(1)Research and implement the file parsing method which can extract MRI image data from the DICOM format.DICOM standard is the medical digital image recognized in the field of image format.This paper analyzes the structure of DICOM format file to design the text information of DICOM file and the analysis method of image data.Extract and store DICOM File's medicine image sequence with C ++ Language,and encapsulates it into a library file so that the medical image segmentation system can call the library file easily.The experiments show that the program can effectively resolve DICOM format files,extract containing patient information and disease data,then display the image.(2)Research and implement medical image segmentation system based on ITK platform..There are most of the medical image processing system for universal platform But different types of images have their own uniqueness,this article focuses on the study MRI medical images.By means of MRI image gray parameters much and sectional imaging characteristics,based on ITK platform,design a image processing system which is mainly used for MRI image segmentation.It has the four modules:image analysis,image preprocessing and segmentation,image display,the design of man-machine interface.Test results show that the system realize the function of human-computer interaction,and can effectively,load DICOM format MRI image,provided by the various algorithms of image segmentation effect is good,clear after the segmentation of the image contour.(3)Implement variety MR! medical image segmentation algorithms arid do performance evaluation.The article uses the medical image segmentation system to achieve variety classic image segmentations and evaluates the reliability,accuracy and efficiency from the subjective aspect and objective aspect and do a lot of experiment to verify.So that according to different types of images the medical staff can choose different algorithms.