In the field of organ transplant matching, UNC has developed algorithms using data-driven mathematical models to predict flow cytometric crossmatch outcomes. If the virtual crossmatch is determined to be likely positive, indicating likely reactivity between recipient and donor, the organ offer is typically declined and issued to another potential recipient. This process is lengthy and delays the overall transplant. These new algorithms project to (1) increase accuracy, from 75-85% to 97% and (2) speed up the process without human bias or intervention.
- Development of data-driven models for the flow cytometric crossmatch Weimer, Eric T., and Katherine A. Newhall. “Development of data-driven models for the flow cytometric crossmatch.” Human immunology 80.12 (2019): 983-989.