“…In contrast to machine learning, CBR does not try to induce rules or models from data: it makes inferences from historical cases directly by comparing them to the current patient [88]. An early example in the field of liver transplantation was presented at AIME 1987 [89]. The CBR community has always been relatively small, but papers on CBR have been presented at many AIME conferences [90][91][92][93].…”