Introducing WFCM into RSF model segmentation of spinal vertebrae

Segmentation of spine vertebrae is an important part of medical image analysis and processing, while also have many challenges and lifting space in this field. In this paper, weighted fuzzy C-means clustering (WFCM) algorithm into RSF(region-scalable fitting) Model segmentation method that can deal with noisy medical images with missing information. While based WFCM to get the rough segmentation images of the CT images, the rough segmentation images are used as the initial contour of the level set evolution. Then using the RSF model to guide the contour curve evolution to the vertebra boundary, getting the final segmentation of vertebrae. The experimental results show that the proposed model can avoid the initialization of the zero curve, which has the characteristics of high accuracy and high efficiency.