Evaluating AI-driven analytics for radiology, part 1: Understanding the lingo and the landscape