A deep-learning-based dose verification tool utilizing fluence maps for a cobalt-60 compensator-based...

Area of Science:

• Medical Physics
• Radiation Oncology
• Artificial Intelligence in Healthcare

Background:

• A novel cobalt-60 compensator-based intensity-modulated radiation therapy (IMRT) system was developed for resource-limited environments.
• This system lacked an efficient algorithm for accurate dose verification.

Purpose of the Study:

• To develop a deep-learning-based dose verification algorithm.
• The goal was to achieve accurate and rapid dose predictions for the novel IMRT system.

Main Methods:

• A deep-learning network was utilized to predict radiation doses.
• The network processed inputs including phantom/patient geometry, beam masks, and fluence maps.
• Two methods were explored for patient-specific dose prediction: field-based and plan-based.

Main Results:

• Deep learning predictions showed high agreement with ground truths for static fields (average deviations <0.5%).
• The plan-based method demonstrated superior agreement for clinical dose distributions compared to the field-based method.
• Dose deviations for target volumes and organs at risk were within 1.3 Gy, with calculations completed in under two seconds per case.

Conclusions:

• A deep-learning-based tool provides accurate and rapid dose verification for a novel cobalt-60 compensator-based IMRT system.
• This approach enhances the feasibility of advanced radiation therapy in resource-limited settings.