This interactive learning lab focuses on improving the consistency and reliability of AI-generated responses in radiology and clinical workflows. Participants will explore how to structure AI outputs using Pydantic models and error feedback loops, ensuring accurate and well-organized data. Additionally, the session introduces DSPy, a user-friendly tool that helps optimize how AI models process and respond to clinical data, reducing the need for manual prompt adjustments. 

 Through hands-on Jupyter notebook exercises, attendees will:  

• Learn how to guide AI models to produce structured, predictable responses. 

• Explore real-world applications, such as extracting key details from radiology reports. 

• Discover how DSPy simplifies AI workflows by automatically improving responses over time. 

• By the end of the session, participants will have practical skills to enhance AI accuracy, streamline workflows, and improve data quality in radiology and healthcare settings. 

Course Requirements and Downloads 
This is a BYOD Session. Participants who wish to participate in the hands-on portion of this lab will be required to bring their own laptop with required software installed. Requirements will be distributed closer to the meeting.

Objectives 

• Explain how structured models improve AI-generated responses for clinical use. 
• Demonstrate how to apply AI optimization techniques using interactive coding exercises. 
• Use hands-on examples to extract structured information from radiology reports.