According to ARRS' American Journal of Roentgenology (AJR), machine learning models applied to presently underutilized imaging features could help construct more reliable criteria for organ allocation and liver transplant eligibility.