Artificial Intelligence in Radiation Therapy

PRELIMS.pdf
Preface
Editor biography
Iori Sumida
List of contributors
CH001.pdf
Chapter 1 Introduction
References
CH002.pdf
Chapter 2 Artificial intelligence and machine learning
2.1 Introduction
2.1.1 Foundations, similarities, and differences
2.1.2 Connection to decision making
2.2 Overview of learning methods
2.2.1 Supervised learning
2.2.2 Unsupervised learning
2.2.3 Semi-supervised learning
2.2.4 Reinforcement learning
2.3 Common algorithms
2.3.1 Gaussian mixture models
2.3.2 Regression and classification algorithms
2.3.3 Decision-tree algorithms
2.3.4 Optimal trees
2.3.5 Neural networks
2.4 Summary
2.5 Acknowledgement
References
CH003.pdf
Chapter 3 Overview of AI applications in radiation therapy
3.1 Opportunities of AI applications in modern radiotherapy workflow
3.2 Summary
References
CH004.pdf
Chapter 4 Introduction to CT/MR simulation in radiotherapy
4.1 Simulation procedure in the radiation therapy process
4.2 Immobilization device for radiation therapy
4.2.1 Systematic error and random error
4.2.2 Reproducibility of patient setup
4.3 Image quality and acquisition time
4.4 Image deformation
4.4.1 Deformable image registration
4.4.2 AI driven image deformation
4.4.3 Practical implementation of AI
References
CH005.pdf
Chapter 5 Organ delineation
5.1 Introduction to organ delineation in radiotherapy
5.1.1 Organ delineation in the radiation therapy process
5.1.2 Impact of delineation accuracies
5.2 Organ delineation methodologies
5.2.1 Automated image segmentation techniques and deep learning applications
5.3 Implementation for clinical diseases: targets and normal structures
5.3.1 Head and neck and brain structures
5.3.2 Thoracic and gastrointestinal structures
5.3.3 Pelvic structures
5.4 Best practice implementation of AI driven delineation
5.5 Future developments and outlook
References
CH006.pdf
Chapter 6 Automated treatment planning
6.1 Goals and motivations of treatment planning
6.2 Automated treatment planning overview
6.3 Knowledge-based planning
6.4 Protocol-based planning
6.5 Multicriteria optimization
References
CH007.pdf
Chapter 7 Artificial intelligence in adaptive radiation therapy
7.1 Introduction
7.1.1 Advantages of ART
7.1.2 Types of ART, current status and challenges
7.1.3 Overview of current workflow of ART and current challenges
7.1.4 AI and AI-assisted technologies for ART
7.2 The role of AI in ART workflow
7.2.1 Deep learning for improving in-room image quality and generating pseudo-CT
7.2.2 Deep learning for deformable image registration and auto-segmentation
7.2.3 Machine learning for decision support on daily adaptation
7.2.4 Machine learning for online re-optimization
7.2.5 AI for quality assurance, verification, and error detection
7.2.6 AI for physics plan check
7.2.7 Considerations for education and training
7.3 Existing AI solutions for ART
7.3.1 Ethos online ART platform from Varian medical
7.3.2 Machine learning solutions from RaySearch Laboratories
7.3.3 PreciseART offline dose monitoring platform from Accuray
7.4 Summary
References
CH008.pdf
Chapter 8 AI-augmented image guidance for radiation therapy delivery
8.1 Introduction to image guidance for radiotherapy
8.1.1 Background
8.1.2 Current image guidance solutions
8.1.3 AI tools and networks for image guidance
8.2 Image guidance for interfraction motion
8.2.1 Patients setup based on orthogonal kV images
8.2.2 Pretreatment daily cone-beam CT imaging
8.3 Image guidance for intrafraction motion
8.3.1 Real-time monitoring methods
8.3.2 Real-time needle and fiducial segmentation
8.4 Real-time 3D IGRT on standard linac
8.5 Summary
References
CH009.pdf
Chapter 9 AI for quality management in radiation therapy
9.1 QA versus QC
9.2 AI for chart review
9.3 AI for patient specific QA and gamma passing rate prediction
9.4 AI for dosimetric and mechanical QA for linear accelerators
9.4.1 Output factor and monitor unit
9.4.2 Linac mechanical error detection
9.5 Summary
References
CH010.pdf
Chapter 10 Data-driven approaches in radiotherapy outcome modeling
10.1 Introduction
10.2 Analytical dose–response models and extensions
10.2.1 Linear-quadratic model and equivalent dose
10.2.2 Tumor control probability and normal tissue complication probability
10.3 Overview of machine learning models
10.3.1 Endpoint prediction: regression and classification
10.3.2 Inclusion of imaging data
10.3.3 Survival prediction models
10.3.4 Performance evaluation metrics
10.4 Practical considerations—building models for radiation oncology
10.4.1 Input data
10.4.2 Feature importance and selection
10.4.3 Tuning hyperparameters
10.4.4 Resampling: cross-validation and bootstrapping
10.4.5 Nested cross-validation and final model selection
10.4.6 Model validation
10.5 Including dose distributions into data-driven outcome models
10.5.1 Voxel-based analysis
10.6 Model reporting: TRIPOD and study analysis plans
10.6.1 Study analysis plans
10.7 Conclusion and future challenges
References
CH011.pdf
Chapter 11 Challenges in artificial intelligence development of radiotherapy
11.1 Radiomics: past, current, and future
11.1.1 Multiparametric radiomics
11.1.2 Multi-radiomics
11.1.3 Artificial intelligence (AI)-empowered radiomics
11.1.4 Precision radiotherapy
11.2 AI and multi-radiomics as a hybrid way for AI development
11.3 Ethics and regulations for artificial intelligence using biomedical informatics
11.4 Heterogeneous biomedical data management
11.5 Human harms due to AI
References