Automation detection of sonography

• Main Author: Koh, Shi Hui
• Other Authors: Zhou Yufeng
• Format: Final Year Project
• Language: English
• Published: 2018
• Subjects: DRNTU::Library and information science::Libraries::Automation
• Online Access: http://hdl.handle.net/10356/75747
• Institution: Nanyang Technological University

Breast cancer is the utmost frequent cancer that happen to the female all over the universe. Although the breast cancer is most common for females at the same time is also the top killer, which causes death to the females. Since the cause of the disease remains unknown, the disease is curable if detected in the early stage. Early detection will increase the success of treatment, giving them a high chance of survival and reducing the cost of treatment. Ultrasound images is the most common use in the world to help to detect and classify them into benign or malignant. In Singapore, breast cancer is the top ranking that causes death to the females. (Hin Peng, et al., 2017) In the past 40 years, the occurrence of breast cancer has twice as large from 25 percent to 65 percent per 100,000 females. In between the year of 2011 and 2015, total up to 2105 females die of breast cancer. In addition, over half of the female were diagnosticate between the age of 45 and 64 years old. Those females diagnose with the early stage, more than 90 percent survive beyond five years as compare to those with 20 percent for those discovered at stage 4. (Khalik, 2017) The accomplished fact conferred above exemplify the jounce of breast cancer. In the past, there is limitation in screening of early stage breast cancer, especially those females in younger age. Ultrasound images are afflicted with bespeckle sonance due to an air gap between the transducer probe and the body. Therefore, pre-processing the images to denoise by using continuous wavelet transform. The aim of this project is to distinguish the cancer in the ultrasound images by using the continuous wavelet transform (CWT) as well as judging the disease by using the artificial neural network (ANN).