• Uses simple image features and whole number cutoffs 
• Straightforward, single-step categorization 
• Provides flexible and useful management recommendations 
• Easy to remember annual follow-up intervals 

Recommendation: Development of RR-1 systems is acceptable  

• Uses workable image features and intuitive decimal number cutoffs
• Simple, two-step categorization process
• Provides mildly prescriptive and practical management recommendations
• Graspable follow-up intervals like 6, 12 and 24 months 

​Recommendation: Development of RR-2 systems is questionable

• Uses arbitrary decimal number cutoffs on things you normally wouldn’t measure
• Moderately complex, multi-step categorization process likely to require a chart
• Provides multifaceted, often obscure management recommendations
• Uses seemingly random follow-up intervals that differ according to patient demographics 

​Recommendation: Development of RR-3 systems is regrettable

• Uses dubious imaging features, non-intuitive oblique measurements in non-standard planes, ratios, percentages and/or enhancement curves
• Highly complex, multi-step categorization process resembling a Rube Goldberg machine
• You’ll be needing a color-coded chart or three 
• Provides costly and invasive management recommendations that keep everyone busy
• Follow-up intervals that ignore the linearity of time and require knowledge of patient risk factors not found in the EMR

​Recommendation: Development of RR-4 systems is unthinkable

• Uses a series of measurements, calculations and qualitative judgements, each requiring its own separate reference article to be at hand 
• Multi-step, multi-dimensional categorization process resembling an Escher drawing
• Requires a wall of color charts to help you plot the conspiracy 
• Provides boisterous management recommendations that can’t be ignored despite their futility 
• Follow-up intervals determined by an online pay-walled black-box calculator 

​Recommendation: Development of RR-5 systems is inexcusable