Computational Image Processing and Deep Learning with Multi-Model Biomedical Image Data Open Access

Abstract

With the rapid advance in medical imaging technology in recent decades, computational image analysis has become a popular research topic in the field of biomedical informatics. Images from various imaging acquisition platforms have been widely used for the early detection, diagnosis, and treatment response assessment in a large number of disease and cancer studies. Although conventional computational methods present higher analysis efficiency and less variability than manual analyses, they require appropriate parameter settings to achieve optimal results. This can be demanding for medical researchers lacking relevant knowledge about computational method development. In the last decade, deep neural networks trained on large-scale labeled datasets have provided a promising and convenient end-to-end solution to biomedical image processing. However, the development of deep-learning tools for biomedical image analysis is often restrained by inadequate data with high-quality annotations in practice. By contrast, a large number of unlabeled biomedical images are generated by daily research and clinical activities. Thus, leveraging unlabeled images with semi-supervised or even unsupervised deep learning approaches has become a significant research direction in biomedical informatics analysis.

   

 My primary doctoral research focuses on the field of medical image processing, utilizing computational methods to facilitate biomedical image analysis with limited supervision. I have explored two ways to achieve this goal: (1) Optimizing the model of existing approaches for specific tasks and (2) Developing semi-supervised/unsupervised deep learning approaches. In my research, I mainly focus on image segmentation and object tracking, two common biomedical image analysis tasks. By experimenting with different types of images (e.g., fluorescence microscopy images and histopathology microscopy images) from various sources (e.g., bacteria, human liver biopsies, and retinal pigment epithelium tissues), my developed methods demonstrate their promising potential to support biomedical image analysis tasks.

Table of Contents

1. Introduction 1

1.1. Research Contributions 2

1.2 Paper list 5

1.3 Outlines 6

2 Literature Review 8

2.1 Deep learning 8

2.2 Image segmentation 10

2.3 Object tracking 11

2.4 Data augmentation 12

3 Object Motion Analysis for Time-Lapse Image Sequences 14

3.1 Method 16

3.1.1 Particle tracking framework 16

3.1.2 Object segmentation 17

3.1.3 Observation and dynamics models 19

3.1.4 Multiple object tracking management 23

3.2 Results 25

3.2.1 Validation with artificial data 26

3.2.2 Bacteria motility analysis 27

3.2.3 Tumor spheroid study 32

3.2.4 Summary 38

4. Biomedical Image Segmentation with Supervised Learning 39

4.1 Method 41

4.1.1 Deep neural network architecture 41

4.1.2 Model implementation 45

4.1.3 Portal tract guided fibrosis quantification 47

4.1.4 Statistical analysis

4.2 Results 48

4.2.1 Training and testing datasets 48

4.2.2 Deep learning model validation 49

4.2.3 Ablation study 57

4.2.4 Clinical correlation analysis 59

4.3 Discussion and summary 62

5. Biomedical Image Segmentation with Semi-supervised Learning 65

5.1 Methods 68

5.1.1 Deep neural network architecture 68

5.1.2 Model implementation 70

5.1.3 Evaluation metrics 73

5.2 Results 75

5.2.1 Training and testing datasets 75

5.2.2 Deep learning model validation 76

5.2.3 Ablation study 79

5.3 Discussion and summary 83

6. Biomedical Image Segmentation with Self-supervised Learning 85

6.1 Methods 88

6.1.1 Deep neural network architecture 88

6.1.2 Image augmentation 90

6.1.3 Model implementation 92

6.2 Results 93

6.2.1 Training and testing datasets 93

6.2.2 Deep learning model validation 94

6.2.3 Ablation study 96

6.3 Discussion and summary 99

7. Conclusion 105

Bibliography 107

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Computer Science and Informatics
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Biology, Bioinformatics Computer Science
Keyword	Semi-supervised learning Image segmentation Self-supervised learning Object tracking Deep learning Biomedical image analysis
Committee Chair / Thesis Advisor	Kong, Jun, Emory University Sunderam, Vaidy , Emory University
Committee Members	Banerjee, Imon , Emory University Nickerson, John, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Computational Image Processing and Deep Learning with Multi-Model Biomedical Image Data ()	2022-11-04 21:34:32 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions