Fetal Cardiac Structure Detection from Ultrasound Sequences

Fetal heart abnormalities are the most common congenital anomalies and are also the leading cause of infant mortality related to birth defects. More than one-third of all malformations found after delivery are congenital heart defects. The prenatal detection of fetal cardiac structure is difficult because of its small size and rapid movements but is important for the early and effective diagnosis of congenital cardiac defects. A novel method is proposed for the detection of fetal cardiac structure from ultrasound sequences. An initial pre-processing is done to remove noise and enhance the images. An effective K means clustering algorithm is applied to the images to segment the region of interest. Finally an active appearance model is proposed to detect the structure of fetal heart. KeywordsCongenital heart defects, Fetal heart, K means clustering, Median filtering, Ultrasound images.