Kim, Sunhee Lee, Deukhee Park, Sehyung Kim, Youngjun 2024-01-19T12:07:36Z 2024-01-19T12:07:36Z 2022-03-01 2014-08 1557-170X https://pubs.kist.re.kr/handle/201004/115334 In this paper, we present a new active contours model to segment human leg bones in computed tomography images that is based on a variable-weighted combination of local and global intensity. This model can split an object surrounded by both weak and strong boundaries, and also distinguish very adjacent objects with those boundaries. The ability of this model is required for segmentation in medical images, e.g., human leg bones, which are usually composed of highly inhomogeneous objects and where the distances among organs are very close. We developed an evolution equation of a level set function whose zero level set represents a contour. An initial contour is automatically obtained by applying a histogram based multi-phase segmentation method. We experimented with computed tomography images from three patients, and demonstrate the efficiency of the proposed method in experimental results. English IEEE Automatic Segmentation of Leg Bones by Using Active Contours Conference 1 36th Annual International Conference of the IEEE-Engineering-in-Medicine-and-Biology-Society (EMBC), pp.4695 - 4698 36th Annual International Conference of the IEEE-Engineering-in-Medicine-and-Biology-Society (EMBC) 4695 4698 US Chicago, IL 2014-08-26 2014 36TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY (EMBC) 000350044704171 2-s2.0-84929467218