Author ORCID Identifier

https://orcid.org/0000-0002-6700-6664

Year of Publication

2020

Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation

College

Engineering

Department

Computer Science

First Advisor

Dr. Nathan Jacobs

Abstract

Medical imaging is an important non-invasive tool for diagnostic and treatment purposes in medical practice. However, interpreting medical images is a time consuming and challenging task. Computer-aided diagnosis (CAD) tools have been used in clinical practice to assist medical practitioners in medical imaging analysis since the 1990s. Most of the current generation of CADs are built on conventional computer vision techniques, such as manually defined feature descriptors. Deep convolutional neural networks (CNNs) provide robust end-to-end methods that can automatically learn feature representations. CNNs are a promising building block of next-generation CADs. However, applying CNNs to medical imaging analysis tasks is challenging. This dissertation addresses three major issues that obstruct utilizing modern deep neural networks on medical image analysis tasks---lack of domain knowledge in architecture design, lack of labeled data in model training, and lack of uncertainty estimation in deep neural networks. We evaluated the proposed methods on six large, clinically-relevant datasets. The result shows that the proposed methods can significantly improve the deep neural network performance on medical imaging analysis tasks.

Digital Object Identifier (DOI)

https://doi.org/10.13023/etd.2020.429

Funding Information

This work was sponsored by Grant No. IIS-1553116 from the U.S. National Science Foundation during 08/2018-05/2020 and Grant No. IRG-19-140-31 from the American Cancer Society during 08/2018-05/2019.

Share

COinS