Speckle Noise Reduction and Segmentation Techniques in Ultrasound for Detection of Fetal Abnormality - A Survey
Ultrasound is a non-invasive diagnostic medical tool used by the healthcare practitioner to evaluate, diagnose and treat medical conditions. Obstetric ultrasonography is an imaging technique used to create real time visual images of the developing fetus in the uterus and is used to detect abnormalities such as heart diseases, kidney problems, limb abnormalities, chromosomal abnormalities, bone disorders, etc. A major problem prevailing in fetal ultrasound is the lack of clarity of an image and is unable to be diagnosed by the physicians. In the ultrasound detection of fetal abnormalities, the edges and the fine details of the images are affected by speckle noise. It is a granular noise that degrades the quality of images, and it affects image interpretation. This can be overcome by using various filtering techniques. The fetal abnormalities can be analyzed and interpreted using image segmentation which involves detection, recognition and measurement of objects in images. The quality of an image is influenced and also enhanced by various image segmentation methods. This paper is a survey of recent studies developed for de-noising and segmentation of ultrasound images in the detection of fetal abnormality. So far the methods are implemented using MATLAB, and for better processing, the filtering and segmentation techniques can be designed and developed using Open-CV in Python.
Obstetric ultrasonography, De-noising, Segmentation, Ultrasound image, Filters.