Research And Implement Of Liver Segmentation Technique Based On Abdomen MRI

MRI is a new iconography method, MRI can show the texture dissection structure clearly, ascertain or remove the most of pathological changes, so the MRI is using extensive in medical treatment diagnoses. However, because of the complexity and uncertainty of the MRI imaging process and the existing abroad otherness among different organizations units, there are important theory meaning and practicality value in automatically and exactly segmenting the interested region in the abdomen MRI scan images.In this paper, I offer a new method to automatically segment the liver in the abdomen MRI scan image. In this method, compress the image data ;using region average and median smoothing method to smooth the image; using Level set method and region-growing method of multi-seed vote mechanism to automatically segment the image; through the automatically segmentation region, error of iconography doctors' manual segmentation region and the average shortest distance to value the effect of automatically segmented image. I use this method in 30 actual abdomen MRI images with fuzzy edges, and prove the validity of the complex image segmentation.