Automatic segmentation of phalanges regions on MR images based on MSGVF snakes

In recent years, medical imaging is important in medical diagnosis for early detection of lesions. However, a large number of images increases the stress to the radiologist. Therefore, CAD (Computer Aided Diagnosis) system is expected to reduce such burden. In diagnostic imaging of phalanges, X-ray photographs, CT (Computed Tomography) are used to evaluate the value of phalanges destruction. Whereby, the MRI (Magnetic Resonance Imaging) that is used mainly to diagnose the lesion in the soft tissue is more effective in a certain case, which is one of the important CAD systems to develop. However, studies on CAD system using MR images are not as much as the studies using CT. In this paper, we propose an automatic segmentation algorithm of phalanges regions on MR images. Although it has three dimensional information, this is the method for two dimensional algorithm. In other words, we propose for each slice of MR Images. Firstly, phalanges regions in MR images are segmented for coarse regions mainly by watershed algorithm. Next, the segmented results from the previous phase are set as the initial contour. Ultimately, the accurate segmentation of the phalanges on MR images are acquired based on MSGVF snakes.