Fully Automated Integrated Segmentation of Carotid Artery Ultrasound Images Using DBSCAN and Affinity Propagation

Abstract Purpose B-mode ultrasound images are used in identifying the presence of fat deposit if any carotid artery. The intima media, lumen, bifurcation boundary is detected by echogenic characteristics embedded Methods A fully automatic self-learning based segmentation proposed extracting edges a modified affinity propagation, which given as inputs to Density Based Spatial Clustering Applications with Noise (DBSCAN) for super pixel segmentation. segmented results analyzed Gradient Vector Flow (GVF) snake model and Particle Swarm Optimization (PSO) clustering using various performance measures. Results parameter free, method combining Affinity propagation DBSCAN evaluated database 361 gives reinforced longitudinal images. approach an improved accuracy 12% increase when compared manual 15% performed individually. average Root Mean Square Error (RMSE) 110 ± 44 µm. Conclusion Extracted edge points automated artery approach.