Automatisierte leitlinienkonforme patientenindividuelle Blutproduktezuordnung und smartes Logistikmanagement in der Transfusionsmedizin

Blood donations are needed as blood products for surgeries and to treat diseases. In view of the declining willingness to donate and the demographic development in Germany, an increasing shortage of donated blood is to be expected in the future.

The aim of the AutoPiLot project is to improve the patient-related use and storage of blood products by systematically using the large amounts of digital health data available in the hospital to make optimized decisions. The project pursues three complementary approaches to increase patient safety, quality of care and cost-effectiveness:

(1) Decision support when ordering blood products by knowledge-based expert systems and intelligent human-machine interaction to increase patient safety

(2) AI-supported patient-specific blood product allocation through the recognition and analysis of patterns and prediction of the patient-specific transfusion requirements to increase patient safety and to optimize warehousing

(3) Prediction of the hospital-wide need for blood products using AI to improve logistics management and reduce the expiration of blood products as well as targeted donor mobilization using a smartphone app.

Institute for AI in Medicine (IKIM)