The Spinal cord analysis plays an important role in neurological disease diagnosis. The segmentation of spinal cord images becomes essential for earlier detection of diseases. At present, radiation therapy is being planned by a radiologist on the basis of a careful study and analysis of a Computed Tomography (CT) scan that covers the tumor and the surrounding tissues. The current planning procedure involves manual description of the spinal cord in each separate slice followed by an automatic reconstruction performed by the image analysis workstation that is connected with the CT-scanning device. Despite the existence of several approaches for planning of repetitive radiotherapy, the detection of the spinal cord in CT images remains an unresolved problem. A factor that complicates the analysis further is the occasional presence of the spine suspended in the spinal canal. Therefore accurate segmentation of the injured spinal cord from CT images can be used. In this paper, Fuzzy C-Means (FCM) clustering is used to identify the vertebrae of the spine. Later Level Set Method is introduced for the edge detection. Then FCM is integrated with Level Set Method involving exact boundary detection to get the proposed method of spatial fuzzy clustering with level set methods (SFCLSM) segmentation. The proposed method gets the accurate results of segmentation of the spinal cord organs. The execution of the proposed method is tested in term of accuracy.
Volume 11 | 02-Special Issue
Pages: 794-801