IntraVascular UltraSound (IVUS) image segmentation is still an open problem as no ap- propriate solution has been discovered yet. The segmentation of IVUS images involves extracting the vessel (outer) and lumen (inner) boundaries from the cross-sectional images. This project applies two widely used image processing techniques separately for the vessel and lumen boundary detection, gradient extraction and adaptive k-means clustering, re- spectively and then employs a well-known contour representation, parametric deformable model to extract both the contours. The contours thus extracted are reconstructed into a 3D data model to assist in visualizing the coronary arteries as a tube-like structure. This helps physicians to visualize and measure the deposition of plaque for appropriate calcium score diagnosis.