Artificial Intelligence in Healthcare and Medicine

✍ Scribed by Kayvan Najarian (editor), Delaram Kahrobaei (editor), Enrique Dominguez (editor), Reza Soroushmehr (editor)

• Publisher: CRC Press
• Year: 2022
• Tongue: English
• Leaves: 301
• Edition: 1
• Category: Library

✦ Synopsis

This book provides a comprehensive overview of the recent developments in clinical decision support systems, precision health, and data science in medicine. The book targets clinical researchers and computational scientists seeking to understand the recent advances of artificial intelligence (AI) in health and medicine. Since AI and its applications are believed to have the potential to revolutionize healthcare and medicine, there is a clear need to explore and investigate the state-of-the-art advancements in the field. This book provides a detailed description of the advancements, challenges, and opportunities of using AI in medical and health applications. Over 10 case studies are included in the book that cover topics related to biomedical image processing, machine learning for healthcare, clinical decision support systems, visualization of high dimensional data, data security and privacy, bioinformatics, and biometrics. The book is intended for clinical researchers and computational scientists seeking to understand the recent advances of AI in health and medicine. Many universities may use the book as a secondary training text. Companies in the healthcare sector can greatly benefit from the case studies covered in the book. Moreover, this book also:

• Provides an overview of the recent developments in clinical decision support systems, precision health, and data science in medicine

• Examines the advancements, challenges, and opportunities of using AI in medical and health applications

• Includes 10 cases for practical application and reference

Kayvan Najarian is a Professor in the Department of Computational Medicine and Bioinformatics, Department of Electrical Engineering and Computer Science, and Department of Emergency Medicine at the University of Michigan, Ann Arbor.

Delaram Kahrobaei is the University Dean for Research at City University of New York (CUNY), a Professor of Computer Science and Mathematics, Queens College CUNY, and the former Chair of Cyber Security, University of York.

Enrique Domínguez is a professor in the Department of Computer Science at the University of Malaga and a member of the Biomedical Research Institute of Malaga.

Reza Soroushmehr is a Research Assistant Professor in the Department of Computational Medicine and Bioinformatics and a member of the Michigan Center for Integrative Research in Critical Care, University of Michigan, Ann Arbor.

✦ Table of Contents

Cover
Half Title
Title Page
Copyright Page
Contents
Editor Biographies
Contributors
Introduction
1. Machine Learning for Disease Classification: A Perspective
1.1 The Groundwork of Machine Learning for Disease Modeling
1.2 The "Big Brother" of Predictions: Supervised Learning
1.3 A Different Language is a Different Vision of Life: Biomedical Data Feature Selection
1.4 Oh Me, Oh My, Omics: Reduction Techniques for Tackling Large Omics Datasets
1.5 Let's Get Ready to Rumble! Training and Testing of Disease Model Predictions
1.6 The Model Rhythm for Your Algorithm
1.7 Seeing Through the Brush: Decisions Trees
1.8 Buttered Up Approach: Kernel Method
1.9 Deep Blue Sea of Predictions: Deep Learning and Neural Networks
1.10 Disease Model to Disease Specialist: Model Interpretability for Healthcare Stakeholders
1.11 The Future is not Something to Predict
References
2. A Review of Automatic Cardiac Segmentation using Deep Learning and Deformable Models
2.1 Introduction
2.2 Background
2.2.1 Medical Background
2.2.2 Image Modalities
2.2.3 Datasets and Challenges
2.2.4 Evaluation Metrics
2.2.4.1 Dice Index
2.2.4.2 Average Perpendicular Distance (APD)
2.2.4.3 Percentage of Good Contours
2.2.4.4 Jaccard Index
2.2.4.5 Sensitivity, Specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV)
2.2.4.6 Hausdorff Distance
2.3 Deep Learning-Based Approaches
2.3.1 Deep Learning Pre-Requisites
2.3.1.1 Artificial Neural Network (ANN)
2.3.1.2 Training the Network
2.3.1.3 Convolutional Neural Networks
2.3.1.4 Convolutional Layer
2.3.1.5 Convolution Operation
2.3.1.6 Pooling Layer
2.3.1.7 Fully Connected Layer
2.3.1.8 Fully Convolutional Neural Networks (FCN)
2.3.1.9 U-Net
2.3.2 Deep Learning for Segmentation of the Cardiovascular Images
2.3.2.1 The Main Deep Learning-Based Cardiovascular Image Segmentation Methods
2.3.2.2 Single-Stage and Multi-Stage Segmentation
2.3.2.3 Structures of Interest
2.3.2.4 Modality
2.3.3 Some Approaches With More Detailed
2.3.3.1 Approach L1
2.3.3.1.1 Overview
2.3.3.1.2 Training/Test
2.3.3.2 Approach L2
2.3.3.2.1 Overview
2.3.3.2.2 Architecture and Algorithm Details
2.3.3.3 Approach L3
2.3.3.3.1 Overview
2.3.3.3.2 Architecture and Algorithm Details
2.3.3.4 Approach L4
2.3.3.4.1 Overview
2.3.3.5 Approach L5
2.3.3.5.1 Overview
2.3.3.5.2 Architecture and Algorithm Details
2.3.3.6 Approach L6
2.3.3.6.1 Overview
2.3.3.6.2 Architecture and Algorithm Details
2.3.3.6.3 Training/Test
2.4 Deformable Models Combined with Deep Learning
2.4.1 Deformable Models Theory
2.4.1.1 Parametric Deformable Models (Active Contours)
2.4.1.2 Geometric Deformable Models (Level Sets)
2.4.1.3 Curve Evolution Theory
2.4.2 Segmentation Approaches using Deformable Models
2.4.2.1 Level-Set and Active Contour
2.4.2.2 Shape Prior
2.4.2.3 2D Deformable Models vs 3D Deformable Models
2.4.3 Combining Deformable Models and Deep Learning
2.4.3.1 Approach H1
2.4.3.2 Approach H2
2.4.3.3 Approach H3
2.5 Conclusion
Notes
References
3. Advances in Artificial Intelligence Applied to Heart Failure
3.1 Introduction
3.2 Understanding Artificial Intelligence
3.2.1 Types of AI
3.2.1.1 Rule-Based Systems
3.2.1.2 Systems Based on Machine Learning
3.2.2 Understanding Machine Learning
3.2.3 The Importance of Data Pre-Processing
3.2.4 Types of Learning
3.2.5 Most Commonly Used AI Techniques for Heart Failure
3.3 Diagnosing Heart Failure
3.3.1 Electrocardiography
3.3.2 Echocardiography
3.3.3 Electronic Health-Related Data (EHRD)
3.4 HF Subtypes Classification and Clustering
3.5 Stratification Risk and Precision Medicine in Heart Failure
3.6 Future Applications and Personalized Medicine
3.7 Conclusions
References
4. A Combination of Dilated Adversarial Convolutional Neural Network and Guided Active Contour Model for Left Ventricle Segmentation
4.1 Introduction
4.2 Literature Review
4.2.1 Image Processing Techniques
4.2.2 Deformable Models
4.2.3 Atlas and Registration
4.2.4 Deep Learning Techniques
4.2.5 Hybrid Techniques
4.2.6 Motivation of the Proposed Method
4.3 Methods
4.3.1 Preprocessing
4.3.2 Convolutional Neural Network
4.3.3 Guided Active Contour Model
4.3.4 Evaluation Metrics
4.4 Experimental Setup
4.4.1 Dataset
4.4.2 Configurations
4.4.3 Evaluation Process
4.5 Results and Discussion
4.5.1 Model Selection
4.5.2 GACM
4.5.3 LV2011 Dataset
4.5.4 Sunnybrook Dataset
4.5.5 Quality Analysis
4.6 Conclusion
References
5. Automated Methods for Vessel Segmentation in X-ray Coronary Angiography and Geometric Modeling of Coronary Angiographic Image Sequences: A Survey
5.1 Introduction
5.2 Pre-Processing Techniques
5.3 Segmentation Methods
5.3.1 Non-Learning Methods
5.3.2 Learning Methods
5.4 Post-Processing Techniques
5.5 Evaluation Metrics
5.6 Applications
5.6.1 Stenosis Measurement
5.6.2 3D Geometric Modeling and Reconstruction
5.7 Summary and Discussion
References
6. Super-Resolution of 3D Magnetic Resonance Images of the Brain
6.1 Introduction
6.2 Super-Resolution: Definition
6.3 Previous Works
6.3.1 Traditional Methods
6.3.2 Deep Neural Networks
6.4 Improved Deep Learning Methods
6.5 Examples and Comparisons
6.5.1 Methods
6.5.2 Datasets
6.5.3 Measures
6.5.4 Results
6.6 Conclusions
References
7. Head CT Analysis for Intracranial Hemorrhage Segmentation
7.1 Introduction
7.2 Hematoma Segmentation
7.2.1 Pre-processing
7.2.1.1 Skull Removal
7.2.1.2 Intensity Adjustment
7.2.1.3 Noise Reduction
7.2.1.4 Image Standardization
7.2.2 Hematoma Segmentation Techniques
7.2.2.1 Classical Image Processing Techniques
7.2.2.2 Deep Learning Techniques
7.2.2.3 Hybrid of Deep Learning and Classical Image Processing Techniques
7.3 Segmentation Performance Evaluation
7.4 Discussion
References
8. Wound Tissue Classification with Convolutional Neural Networks
8.1 Introduction
8.1.1 Wound Diagnosis
8.1.2 Workflow for Wound Image Evaluation
8.2 Related Works
8.2.1 Machine Learning for Wound Image Evaluation
8.2.2 Deep Learning Models for Wound Image Evaluation
8.3 Methodology
8.3.1 ROIs Generation
8.3.2 ROIs Classification
8.3.3 Segmentation Process
8.4 Experimental Results
8.5 Conclusions
References
9. Artificial Intelligence Methodologies in Dentistry
9.1 Introduction
9.2 AI Techniques in Dentistry
9.2.1 AI Techniques using Imaging Data
9.2.2 AI Techniques using Integrative Multi-source Data Models
9.3 Evaluation Metrics
9.4 Final Considerations
References
10. Literature Review of Computer Tools for the Visually Impaired: A Focus on Search Engines
10.1 Introduction
10.2 Research Methods
10.3 Tools
10.3.1 What Are Tools?
10.3.2 Search Engines (SEs)
10.4 Primary Attributes of Tools
10.4.1 Attribute: Navigation
10.4.1.1 Scope
10.4.1.2 Difficulties Faced by VI Users
10.4.1.3 Existing Products and Ongoing Research
10.4.1.4 Necessary Future Research and Consideration
10.4.2 Attribute: User Interface
10.4.2.1 Scope
10.4.2.2 Difficulties Faced by VI Users
10.4.2.3 Existing Products and Ongoing Research
10.4.2.4 Necessary Future Research and Consideration
10.4.3 Attribute: Information Accessibility
10.4.3.1 Scope
10.4.3.2 Difficulties Faced by VI Users
10.4.3.3 Existing Products and Ongoing Research
10.4.3.4 Necessary Future Research and Consideration
10.4.4 Attribute: Latency
10.4.4.1 Scope
10.4.4.2 Difficulties Faced by VI Users
10.4.4.3 Existing Products and Ongoing Research
10.4.4.4 Necessary Future Research and Consideration
10.4.5 Attribute: Discreetness
10.4.5.1 Scope
10.4.5.2 Difficulties Faced by VI Users
10.4.5.3 Existing Products and Ongoing Research
10.4.5.4 Necessary Future Research and Consideration
10.4.6 Attribute: Emotional Implications
10.4.6.1 Scope
10.4.6.2 Difficulties Faced by VI Users
10.4.6.3 Existing Products and Ongoing Research
10.4.6.4 Necessary Future Research and Consideration
10.5 Understanding the Userspace
10.6 Evaluations and Surveys
10.6.1 Evaluations
10.6.2 Visual Question Answering Surveys
10.7 Understanding the Big Picture
Notes
References
11. Tensor Methods for Clinical Informatics
11.1 Introduction
11.2 Definitions and Notation
11.3 Matrix and Tensor Products
11.4 Tensor Decompositions
11.4.1 CP Decomposition
11.4.2 Tucker Decomposition
11.4.3 Other Decompositions
11.5 Tensor Completion
11.6 Tensor Decomposition Applications - A Few Examples
11.6.1 Brain Data
11.6.2 Heart Data
11.6.3 EHR Data
11.6.4 Drug Repurposing
11.6.5 Other Applications
11.7 Conclusion
References
Index