MR-linac: role of artificial intelligence and automation

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Artificial Intelligence in Medicine

Background:

• The past five years have seen significant AI integration in radiotherapy, automating organ at risk delineation and treatment planning.
• Magnetic resonance (MR)-guided linear accelerators (MR-linacs) have improved treatment accuracy and real-time adaptation, especially for tumors near sensitive organs.

Purpose of the Study:

• To highlight the advancements and ongoing needs for AI in radiotherapy, particularly for MR-guided radiotherapy (MRgRT).
• To explore how AI can streamline MRgRT workflows, reduce manual processes, and improve patient outcomes.

Main Methods:

• Review of recent advancements in AI for radiotherapy automation.
• Discussion of AI applications in synthetic CT generation, automated contouring, and treatment planning.
• Exploration of AI in quality assurance, motion management, and outcome prediction.

Main Results:

• AI automation enhances consistency, accuracy, and efficiency in radiotherapy processes.
• AI integration in MR-guided radiotherapy (MRgRT) promises to reduce labor intensity and treatment times.
• AI-driven quality assurance and predictive biomarkers are crucial for patient safety and treatment effectiveness.

Conclusions:

• AI is essential for optimizing MR-guided radiotherapy (MRgRT) workflows and expanding access to advanced technologies.
• Further AI development will improve treatment precision, patient safety, and enable personalized treatment strategies through outcome prediction.