Date of Award

Spring 2021

Document Type

Thesis

Publisher

Santa Clara : Santa Clara University, 2021.

Department

Bioengineering

First Advisor

Yuling Yan

Abstract

Breast cancer is the second leading cause of cancer deaths among US women. Thus, it is important for doctors to detect and diagnose breast cancer as early as possible. Mammography has been used for about 30 years, but there have been rapid developments using digital mammography technology and computer aided systems to help improve breast imaging. Deep learning techniques are being developed to provide a more effective tool for the classification of breast cancer. We adopt a transfer learning approach and fine-tune a pre-trained convolutional neural network model for accurate classification of breast masses based on screening mammograms. The model is retrained and tested using the CBIS-DDSM (Curated Breast Imaging Subset of Digital Database for Screening Mammography) dataset. We are able to achieve a training accuracy of 71.1% and a test accuracy of 68.7%.

Share

COinS