Machine Learning in Dentistry

Preface
Machine Learning in Dentistry
Machine Learning for Dental Imaging
Machine Learning for Oral Diagnosis and Treatment Planning
Machine Learning and Dental Designs/Outcome Assessment
Machine Learning Supporting Dental Research
Contents
Part I Machine Learning for Dental Imaging
1 Machine Learning for CBCT Segmentation of Craniomaxillofacial Bony Structures
1.1 Introduction
1.2 Background
1.2.1 Random Forest
1.2.1.1 Construction of a Binary Decision Tree
1.2.1.2 Construction of a Random Forest
1.2.2 Feature Extraction
1.3 Prior-Guided Sequential Random Forests
1.3.1 Training
1.3.1.1 Initial Probability Maps
1.3.1.2 Complementary Imaging Features
1.3.1.3 Sequential RFs
1.3.2 Inference
1.4 Experiments
1.4.1 Dataset and Pre-processing
1.4.2 Experimental Setting
1.4.3 Segmentation Results of Maxilla and Mandible
1.4.4 Segmentation Results of Teeth
1.4.5 Importance of Prior and Sequential Learning
1.5 Conclusion
References
2 Machine Learning for Craniomaxillofacial Landmark Digitization of 3D Imaging
2.1 Background
2.2 Related Methods for Landmark Localization
2.3 Craniomaxillofacial Landmark Digitization
2.3.1 Multi-Atlas-Based Landmark Digitization
2.3.2 Regression Forest-Based Landmark Digitization
2.3.3 Landmark Digitization Using Segmentation-Guided Partially Joint Regression Forest
2.3.4 Joint Bone Segmentation and Landmark Digitization Using Deep Learning
2.4 Results
2.5 Conclusion
References
3 Segmenting Bones from Brain MRI via Generative Adversarial Learning
3.1 Introduction
3.2 Related Works
3.2.1 Semi-supervised Learning
3.2.2 Cross-Modality Image Synthesis
3.3 Method
3.3.1 Overview
3.3.2 Cross-Modality Image Synthesis for Segmentation
3.3.3 Neighbor-Based Anchoring Method
3.3.4 Feature-Matching-Based Semantic Consistency
3.3.5 Network Architecture
3.3.5.1 Generators
3.3.5.2 Discriminators
3.3.5.3 Segmentor
3.3.6 Implementation Details
3.4 Experiment and Results
3.4.1 Experimental Setting
3.4.2 Image Synthesis Results
3.4.3 Segmentation Performance
3.4.4 Ablation Study
3.5 Conclusion and Discussion
References
4 Sparse Dictionary Learning for 3D Craniomaxillofacial Skeleton Estimation Based on 2D Face Photographs
4.1 Introduction
4.2 Method
4.2.1 Patient's Normal 3D Face Reconstruction
4.2.2 Sparse Dictionary Learning for Initial Reference Bone Model Estimation
4.2.3 Refinement of the Initially Estimated Reference Model
4.3 Experiments and Results
4.3.1 Experimental Dataset
4.3.2 Evaluation on Synthetic Dataset
4.3.3 Evaluation on Real Dataset
4.4 Discussion and Conclusion
References
5 Machine Learning for Facial Recognition in Orthodontics
5.1 Introduction
5.2 Background Knowledge in Cranio-facial Recognition
5.2.1 Facial Perception and Recognition by the Brain
5.2.2 Pattern Recognition
5.2.3 Multi-label Image Classification
5.2.4 Convolutional and Recurrent Neural Networks
5.3 Systems that Automatically Provide Clinical Descriptions of Facial Images for Orthodontic Diagnostic Purposes
5.3.1 Sample Data
5.3.2 Model Multi-label Image Classification
5.3.3 Results
5.4 Deep Learning-Based Cephalometric Landmark Identification Using Landmark-Dependent Multi-scale Patches
5.4.1 Dataset and Method
5.4.1.1 Training Phase
5.4.1.2 Landmarking Phase
5.4.2 Evaluations
5.5 Results
5.6 Discussions for Recognition of Medical Facial Images
5.7 Conclusions
References
Part II Machine Learning for Oral Diagnosis and Treatment
6 Machine/Deep Learning for Performing Orthodontic Diagnoses and Treatment Planning
6.1 Introduction
6.2 Various AI Systems for Determining Treatment Plans in Orthodontics
6.3 Overview of Our Automated Orthodontic Diagnosis System
6.4 A System that Automatically Designs an Orthodontic Treatment Plan Based on a Document of Medical Findings Describing a Patient's Condition
6.4.1 Dataset and Problem Setting
6.4.2 Medical Findings Summarization Task
6.4.2.1 Methods
6.4.2.2 Experiments
6.4.3 Treatment Planning Task
6.4.3.1 Methods
6.4.3.2 Experiments
6.5 Conclusion
References
7 Machine Learning in Orthodontics: A New Approach to the Extraction Decision
7.1 Introduction
7.2 Logistic Regression
7.3 Machine Learning
7.3.1 Three-Layer Artificial Neural Network
7.3.2 Comprehensive Survey
7.3.2.1 Classical and Regression Trees
7.3.2.2 Random Forest
7.3.2.3 Multilayer Perceptron
7.3.2.4 Methods
7.3.2.5 Results and Discussion
7.4 Conclusion
References
8 Machine (Deep) Learning for Characterization of Craniofacial Variations
8.1 Introduction
8.2 Automatic Three-Dimensional Image Segmentation
8.2.1 Preparing Training Samples and Ground Truth
8.2.2 A Machine Learning-Based Method: 3D U-Net
8.2.3 Segmentation Results of Maxillae and Palatal Defect Based on 3D UNet
8.3 Application Review
8.3.1 Unilateral Impacted Canine-Related Maxilla Morphology Variation
8.3.2 Discussions
8.4 Conclusion
References
9 Patient-Specific Reference Model for Planning Orthognathic Surgery
9.1 Introduction
9.2 Method
9.3 Accuracy Evaluation
9.3.1 Data Acquisition
9.3.2 Parameter Setting
9.3.3 Qualitative Evaluation
9.3.4 Quantitative Evaluation
9.4 Discussion
References
Part III Machine Learning and Dental Designs
10 Machine (Deep) Learning for Orthodontic CAD/CAM Technologies
10.1 Introduction
10.2 Background Knowledge in Deep Learning
10.2.1 Loss (Cost) Function and Metrics
10.2.2 Training, Validation, and Test Datasets
10.2.3 Epoch and Batch Size
10.3 Automatic Tooth Segmentation Based on MeshSegNet
10.3.1 PreProcessing
10.3.1.1 Simplification
10.3.1.2 Data Annotation
10.3.1.3 Data Augmentation
10.3.2 MeshSegNet
10.3.3 Segmentation Result
10.3.3.1 Segmentation Result Comparison
10.3.4 Post-Processing
10.3.4.1 Label Refinement
10.3.4.2 Label Mapping
10.4 Conclusion
References
11 Assessment of Outcomes by Using Machine Learning
11.1 Introduction
11.2 Overview of Machine Learning
11.3 Craniofacial Genomics Using Machine Learning
11.4 Machine Learning Tools
11.5 Assessment of Images and Related Outcomes Using Machine Learning
11.5.1 2D Lateral Cephalometric Radiographs
11.5.2 3D Cephalometric Radiographs
11.5.3 CBCT and CT Auto-Segmentation
11.5.4 Digital 3D Dental Models
11.5.5 Remote Dental Monitoring
11.6 Conclusions
References
Part IV Machine Learning Supporting Dental Research
12 Machine Learning in Evidence Synthesis Research
12.1 What Are Systematic Reviews?
12.2 Research Question
12.3 Search
12.4 Study Selection/Screening
12.5 Data Extraction
12.6 Appraisal: Risk of Bias Assessment
12.7 Implementation Challenges/Barriers to Adoption
References
13 Machine Learning and Deep Learning in Genetics and Genomics
13.1 Fundamentals of Machine Learning and Oral Health
13.2 Machine Learning `Omics Applications
13.3 Multi-Omics and Integrative Analyses
13.4 Multi-Omics in Applications in Dental Caries and Early Childhood Oral Health
13.5 Statistical Methods Including Machine/Deep Learning in Association Analysis Between SNPs and Complex Diseases
13.6 ML and DL in CNV Calling
13.7 ML and DL Methods for DNA Methylation Data
13.8 ML and DL Methods for Hi-C Data
13.9 ML and DL Methods in Analysis of Transcription Factors
13.10 ML and DL Methods for Single-Cell Transcriptomics Data
13.11 Machine and Deep Learning in Genomics-Based Oral Health: Systemic Review of Literature
13.11.1 Methods
13.11.2 Results
13.12 Discussion of Limitations About Using ML and DL in Genetics and Genomics in Oral Health
13.13 Summary and Conclusion of ML and DL in Genetics and Genomics in Oral Health
References
14 Machine (Deep) Learning and Finite Element Modeling
14.1 Introduction
14.2 Machine Learning for Geometry in Finite Element Analysis
14.3 Using Machine Learning to Overcome the Computational Expenses in Finite Element Analysis
14.4 Using Finite Element Analysis Results as One Feature in Machine Learning for the Classification
14.5 Discussion and Conclusion
References