Radiologists are beginning to use AI-based computer vision models to help speed up the laborious process of parsing medical scans. However, these models require large amounts of carefully labeled training data to achieve consistent and accurate results, meaning radiologists must still dedicate significant time to annotating medical images.