Robust segmentation and intelligent decision system for cerebrovascular disease.

摘要: Segmentation and classification of low-quality noisy ultrasound images is challenging task. In this paper, a new approach proposed for robust segmentation carotid artery consequently, detecting cerebrovascular disease. The technique consists two phases, in first phase; it refines the class labels selected by user using expectation maximization algorithm. Genetic algorithm then employed to select discriminative features based on moments gray-level histogram. refined targets are fed as input neuro-fuzzy classifier performing segmentation. Finally, intima-media thickness values measured from segmented segregate normal abnormal subjects. second phase, an intelligent decision-making system support vector machine developed utilize (RSC-US) has been tested dataset 300 real yields accuracy, F-measure, MCC scores 98.84, 0.988, 0.9767 %, respectively, jackknife test. performance also at several noise levels may be used secondary observation.