Transfusion Medicine: From AB0 to AI (Artificial Intelligence)

ABSTRACT

Transfusion of blood from human to human became a clinical reality around the turn of the 19^th century with the discovery of the blood group antigens A and B, which allowed compatibility between the blood of the healthy donor and the recipient. During the past century the need for accurate documentation and archiving of blood group and compatibility data became an increasing issue. World War II provided a specific boost in the development with the introduction of the first primitive mechanical data processing allowing a more secure traceability and retrievability of data. Simultaneous development of better collection and preservation technologies allowed banking and transportation of blood. The introduction of machine learning in automation of equipment contributed to uniformity and standardization, followed by the development of deep learning algorithms to predict transfusion needs and compatibility, contributing to improved safety of patient care.