Skip to Main Content
Echocardiogram is one of the effective diagnostics tool for cardiac investigations. Artifacts such as speckle, grating lobes, and shadowing in ultrasound images can hamper experts interpretation and impede automated analysis. This is because noise and artifacts cause edges to manifest themselves in different ways from the typical definition; hence they pose challenge to conventional edge detection and noise suppression methods. Thus, a method is proposed to resolve the ambiguous edge definitions in noisy echocardiogram by applying the adaptive neuro-fuzzy inference system (ANFIS) to determine edgeness based on local image characteristics that are defined by operators using local statistics. The performance of the proposed method is compared to that of a conventional edge detector and another similar method but without learning capability.