Artificial Intelligence in Kidney Transplantation: A Scoping Review (Preprint)

Abstract

BACKGROUND Artificial Intelligence technologies and big data have been increasingly used to enhance kidney transplant experts’ ability to make critical decisions and manage the care plan for their patients. OBJECTIVE To explore the use of AI technologies in the field of kidney transplantation as reported in the literature. METHODS Embase, CINAHL, PubMed and Google Scholar were used in the search. Backward reference list checking of included studies was also conducted. Study selection and data extraction was done independently by three reviewers. Data extracted was synthesized in a narrative approach. RESULTS Of 505 citations retrieved from the databases, 33 unique studies are included in this review. Artificial intelligence (AI) technologies in the included studies were used to help with diagnosis (n= 16), used as a prediction tool (n=15) and, also for supporting appropriate prescription for kidney transplant patients (n = 2). The population who benefited from the technique included patients who underwent kidney transplantation procedure (n = 24) and those who are potential candidate (n=6). The most prominent AI branch used in kidney transplantation care was machine learning with Random Forest (n=11) being the most used AI model, followed by Linear Regression (n=6). CONCLUSIONS Conclusion: AI is extensively being used in the field of kidney transplant. However, there is a gap in research on the limitation and obstacles associated with implementing AI technologies in kidney transplant. There is a need for more research to identify educational needs and standardized practice for clinicians who wish to apply AI technologies in critical transplantation-related decisions.