The trouble is, most of our understanding of “abnormal” imaging findings comes from horses who were imaged because they were lame. When you only ever scan the horses with problems, you start to believe that anything you find must be the cause of those problems.