Design and Analysis of Breast Cancer Detection System using Ultrasound Imaging and Deep Learning

Please enable javascript in your browser.

Shams, Morteza | 2023

0 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 57088 (05)
University: Sharif University of Technology
Department: Electrical Engineering
Advisor(s): Fakharzadeh, Mohammad; Kavehvash, Zahra
Abstract:
One of the most important factors that make breast cancer such a dangerous and deadly disease is its hidden. One of the most important methods of revealing this disease is imaging. There are different types of imaging to reveal cancerous tissues (tumors), such as ultrasound, mammography, X-ray, etc. Ultrasound imaging is a cheap, convenient and accessible imaging. The aim of this project is to provide a breast cancer detection system that uses the images taken with this method according to the characteristics of ultrasonic imaging. Regarding cancer diagnosis, doctors identify the tumor with the help of their personal experience, which is based on the characteristics of the tumor image. In this project, we detect breast cancer by using deep learning and finding characteristics of benign, malignant, and healthy tumor images. This project includes four main parts of pre-processing of ultrasonic images (preparation of images for use), production of mask images (segmented image of tissue and tumor), data augmentation (due to lack of input images) and finally implementation of deep learning model for detection. is
Keywords:
Sonography ; Breast Cancer ; Deep Learning ; Image Processing ; Artificial Intelligence ; Medical Ultrasound Imaging ; Mammography

No TOC