A fine-tuned large language model chatbot for multi-scenario radiology cancer care: randomized controlled trial on interaction optimization, emotional support, and provider burnout reduction

Key Points

• This research aims to evaluate the effectiveness of a chatbot in optimizing interactions and reducing provider burnout in cancer care.
• Conducted a randomized controlled trial
• Implemented a fine-tuned large language model chatbot
• Assessed interactions related to emotional support
• Monitored provider burnout levels
• Chatbot intervention improved emotional support metrics
• Provider burnout levels were significantly reduced
• Enhanced interaction optimization was noted across multiple scenarios