"In 1970 I realized that there was negligible risk from x-rays but many radiographs had poor image quality so that the risk from a false negative was significant."