Research On Medical Image Segmentation Based On Fuzzy Clustering Algorithm

As medical imaging technology, e.g. MRI, X-CT, PET and SPECT, is widely used inhospital, computer aided diagnosis system becomes more and more important in clinicalapplication. Medical imaging play an important role in disease, treatment planning andoperation effect evaluation, for it provides useful information for medical personnel.Magnetic Resonance Imaging can give sound result from any angle and parameter withoutcausing any harm to patient, and because of these advantages, it is especially importantamong all the imaging technologies.Medical image processing is the most important part of the computer aided diagnosissystem and achieve great progress in the past years. Researchers have developed lots ofalgorithm to solve the problem. Fuzzy cluster method is a kind of soft segmentationmethod and it provides sound result for medical image segmentation. Because the Fuzzy CMeans method considers only the gray value information, it is not robust to the noise. Inthis paper, an improved FCM merged with spatial filter is proposed to solve this problem.In our method, the MR image data is denoised with improved spatial filter. Then the objectfunction and the iteration formulas is modified by adding the spatial filter data item so thatthe iteration is affected by the spatial filter. Experimental result shows that the algorithmobtains sound segmentation result of MR images.