Ultrasound Image Analysis for Normal Pregnancy and Ectopic Pregnancy

摘要

Ectopic pregnancy is a complicated state of pregnancy where the embryo engaged outside from the uterus. With infrequent exceptions, the fetus will fail to survive. Detection of ectopic pregnancy usually accomplished by ultrasound and human chorionic gonadotropin (hCG) test in blood. Over two decades, ultrasound imaging exhibits very high safety records. Analyzing ultrasound images is a vital problem for all image processing systems. Also, the ultrasound image consists of a significant amount of noise, such as speckle noise. Moreover, the existence of this noise makes problems in image analysis. Our focus is to eliminate the detection problem by applying some image processing methods. The method includes image filtering, segmentation, and morphological operations. The results show that by the combination of thresholding and morphological operation with the region of interest method, the border of the gestational sac, sagittal ring, and the area of the uterus can be detected. As a result, ectopic pregnancy can be easily distinguished.