A novel deep-learning approach for monitoring gastrointestinal air variation during radiotherapy in young patients using radiographs

Key Points

• This research aims to utilize deep learning to monitor gastrointestinal air changes during radiotherapy in young patients.
• Developed a deep learning-powered system to analyze radiographs for gastrointestinal gas variation.
• Tested across diverse age groups and demographics to ensure generalizability and performance.
• Proposed a personalized image-guided radiotherapy workflow to optimize treatment efficiency.
• Successful quantification of internal gastrointestinal gas changes using deep learning methods.
• Promising performance metrics indicating the potential reduction of CBCT utilization in required treatment fractions.