Level Set Method for Segmentation of Medical Images Without Reinitialization

Accuracy and Clarity are important issues for medical imaging and same in the case of segmentation. This paper proposes a novel method for segmentation of medical image based on level set without reinitialization approach, applied with specific shape model. Level set method without reinitialization, with certain specific shape based model has advantages over level set method with reinitialization. Large time steps are possible with the proposed method which speed up the process of curve evolution. We have applied the proposed method on several medical images. Results on six set of two different images are being presented in this paper. The proposed approach is compared with four recent state-of-art segmentation methods, and the experiment results obtained validate that the proposed method improves the segmentation accuracy, clarity and efficiency for medical images than other methods qualitatively and quantitatively in terms of Mean Square Error (MSE), Peak Signal to Noise Ratio (PSNR), Maximum Difference (MD) and Structural Content (SC).