Loading report data…

// Anthropic API · Pre-generated summaries

Radiology Report Summarizer

This demo showcases outputs from running synthetic radiology reports through the Anthropic API (offline, via a local script)—each report is expanded into four audience-specific summaries: an impression-style summary, key findings, follow-up recommendations, and a plain-language patient version. The live site only loads static JSON; nothing is sent to Anthropic from your browser.

Original report


      


      

        

          
Impression summary

          

        

        

          
Key findings

          
    
        
    
        
    
          
    Follow-up recommendations
    
          
    
        
    
        
    
          
    Patient summary
    
          
    
        
    
      
    
    
    

    
    
      
    // Technical summary
    
      
    What this demo uses
    
      
    
        JavaScript
        Static JSON
        Content design
        Python (offline)
      
    
      
    
        
    
          This demo shows pre-generated report summaries for different audiences, designed to make the same medical content easier to understand at different reading levels.
        
    
        
      
          
    • Methodology: Synthetic reports are summarized offline, saved as static JSON, and then rendered by the site. The browser only reads saved outputs; no live model calls happen in the page.
      • 
          
    • 
            Technical terms:
            
        
              
      • LLM: Large Language Model, an AI system that generates text.
        • 
              
      • Prompt: Instructions given to the model.
        • 
              
      • Static JSON: A pre-saved data file served as-is.
        • 
            
      
          
      • 
          
    • Toolsets/technology used: JavaScript, static JSON rendering, and offline Python scripts with the Anthropic API.
      •