Artificial Intelligence in Proton Therapy: What Works, What Does Not, and What Is Next

Read full text Summarise with VIRAT
Authors: Heng Li, Ming Yang, Doris Keziah Ndassi, Sizhuo Meng, X. Jia
Cited by: 0
References: 62

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Artificial intelligence is increasingly shaping the evolution of proton therapy, with applications spanning imaging, treatment planning, quality assurance, adaptive workflows, and outcome modeling. Unlike conventional task-specific algorithms, modern AI methods, including machine learning and deep learning, enable integration of heterogeneous data and capture complex relationships across the clinical workflow. These capabilities are particularly relevant in proton therapy, where sensitivity to range uncertainty, anatomic variation, and biological heterogeneity presents persistent clinical and operational challenges. This review summarizes current and emerging AI applications in proton therapy, including image reconstruction and synthesis, segmentation, dose prediction, robustness and uncertainty management, biological optimization, and adaptive treatment strategies. We also discuss the expanding role of AI in quality assurance and workflow coordination, emphasizing the distinction between task-level automation and workflow-level intelligence. Finally, we address broader considerations related to clinical validation, safety, interpretability, economic value, and access, which will be critical for translating AI-enabled proton therapy into routine clinical practice.

Fields of Study & Subjects

Radiation Therapy and Dosimetry Artificial Intelligence in Healthcare and Education Advanced Radiotherapy Techniques

Publication details

Journal / source
The Cancer Journal
Volume
32
Issue
2
Publication date
1 March 2026
Language
EN