Ahmed Reda Mostafa Ahmed Elazizy <br/><br/>
Morphological characterization of breast tumors using conventional b-mode ultrasound images /  
 التحليل الوصفى لأورام الثدى باستخدام الصور التقليدية لأجهزة الموجات فوق الصوتية 
Ahmed Reda Mostafa Ahmed Elazizy ; Supervised Ahmed M. Ehab Mahmoud , Muhammed Ali Rushdi 
 - Cairo :   Ahmed Reda Mostafa Ahmed Elazizy ,   2020 
 - 62 P. :   charts , facsimiles ;   30cm  
<br/><br/>Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Systems and Biomedical Engineering  
<br/><br/> This work aims to develop and test a vendor-independent computer-aided diagnosis (CAD) system that uses conventional B-mode ultrasound images to distinguish between benign and malignant breast tumors.Three morphological features were extracted from 323 breast tumor lesions including the perimeter, regularity variance, and circularity range ratio. Lesions were segmented using the active contour method via semi- and fully-automated algorithms.Then, the support vector machine classifier was used to identify breast lesions. Results of the CAD system exhibited accuracies of 95.98% and 95.67% using the semi- and fully-automated segmentation, respectively. Based on the preliminary results, this CAD system with such unique combination of geometrical features shall improve the diagnostic decisions and may reduce the need of unnecessary needle biopsies 
<br/><br/>
<br/><br/><label>Subjects--Index Terms: </label>Breast  Computer-aided diagnosis (CAD)  Ultrasound (US)