Abstract
Liver transplantation is the ultimate therapeutic option for patients with end-stage liver disease. The clinical management of transplant patients significantly impacts their prognosis, with outcomes influenced by multiple interacting variables. Cardiovascular complications count as a leading cause of both short-term and long-term morbidity and mortality in liver transplant recipients. In this respect, accurate risk assessment and stratification are crucial for optimizing clinical outcomes. Modern artificial intelligence (AI) techniques have significant potential for early risk prediction, providing comprehensive risk assessments in both diagnosed cohorts and early clinical phase patients. This systematic review examines the state of the art in AI applications for predicting cardiovascular risks in liver transplantation, identifying current issues, challenges, and future research directions. We reviewed articles from digital repositories such as PubMed, IEEE Xplore, and ScienceDirect published between 2000 and 2023, using keywords including artificial intelligence, machine learning, cardiovascular, and liver transplantation. Our analysis revealed a diverse range of machine learning algorithms used in this domain. Despite the potential, only 12 papers met the criteria for adequate topic coverage, highlighting a scarcity of research at this intersection. Key challenges include integrating diverse datasets, isolating cardiovascular effects amid multifaceted influences, ensuring data quality and quantity, and the issues to extrapolate machine learning models to day-to-day clinical practice. Nevertheless, leveraging AI for risk prediction in liver transplantation could significantly enhance patient management and resource optimization, indicating a shift towards more personalized and effective medical practices.