Abstract:
The main fortitude of diagnostic ultrasound are its ability to reveal the mechanical compositionofanatomy, the dynamic mobility of organs and instant details of bloodcirculation. Breast ultrasoundimaging is congenital, radiation-free, and convenient. Automatic breast ultrasound (BUS) imagesegmentation is difficult in clinicalpractice but vital for cancer diagnosis and medication planning. Speckle noises and attenuated artifacts are prominent in ultrasound images. Speckles obscureaspects that are essential for diagnosis and evaluation. Due to low contrast, automatically findingareas of interest (ROIs) and inaccurate tumor boundaries in breast ultrasound images (BUS), automatically segmenting breast tumors from ultrasound remains a challenge. The majorityofultrasound image segmentation algorithms are centered on region growing or active contours. These are semi- automatic segmenting methods in which seed points or preliminary contours haveto be identified manually. This thesis presents a fully automated tumor segmentationfromUltrasound Images by choosing dynamic thresholding. This framework also includes selectionof anautomated threshold value for watershed method. The area and boundary error metrics are usedtoevaluate the performance of the proposed completely automatic segmentation method on a BUSdatabase. Our proposed method is more accurate and reliable in the segmentation of BUSimagesthan the fully automated method recently proposed.
Description:
Supervised by
Prof. Dr. Md. Taslim Reza,
Department of Electrical and Electronic Engineering(EEE),
Islamic University of Technology(IUT),
Board Bazar, Gazipur-1704, Bangladesh
This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Electrical and Electronic Engineering, 2022.
