Assessing fracture detection: A comparison of minimal-resource and standard-resource plain radiographic interpretations