Artificial intelligence in radiotherapy: Current applications and future trends

Area of Science:

• Medical Physics
• Radiation Oncology
• Artificial Intelligence

Background:

• Computed tomography and intensity modulation have advanced radiation therapy, increasing treatment precision but also workflow complexity.
• Accurate and uniform delineation of numerous volumes became critical with these technological advancements.
• Improved computing power enabled reverse planning and 3D dose distribution generation.

Purpose of the Study:

• To explore the role of artificial intelligence in enhancing radiation therapy workflows.
• To investigate AI's potential in increasing practice homogeneity and efficiency.
• To discuss the development of predictive tools for personalized radiation oncology.

Main Methods:

• Implementation of artificial intelligence-based tools in routine radiation therapy practice.
• Integration of workflow data with clinical and omic data.
• Development of predictive models for clinical decision-making.

Main Results:

• AI tools are increasing efficiency, reducing workload, and improving treatment homogeneity.
• Data integration is enabling the development of proof-of-concept predictive tools.
• These tools show potential for personalized oncologic strategies and dose prescriptions.

Conclusions:

• Artificial intelligence offers significant opportunities to optimize radiation therapy workflows and enhance treatment delivery.
• Predictive tools, though in early stages, are crucial for advancing towards personalized radiation oncology.
• Prospective validation on large, multicenter cohorts is necessary for widespread adoption of AI-driven predictive models.