# Artificial Intelligence in Renal Transplantation Over the Past Decade: A Narrative Review of Clinical Applications, Current Limitations, and Future Directions

## TL;DR

This paper reviews how AI has been used in kidney transplants over the last decade, focusing on improvements in donor matching, surgery, and outcome predictions.

## Contribution

The paper provides a comprehensive narrative review of AI applications in kidney transplantation, emphasizing recent clinical advancements and limitations.

## Key findings

- AI improves donor-recipient matching and survival prediction in pretransplant settings.
- AI enhances robotic surgery with augmented reality and 3D models for better preoperative planning.
- Artificial neural networks outperform traditional methods in predicting post-transplant graft survival and rejection.

## Abstract

This narrative review examines the use of artificial intelligence (AI) and machine learning (ML) in kidney transplantation (KT) during the past 10 years, highlighting advancements in clinical applications and future potential. In pretransplant settings, AI algorithms assist in matching donors with recipients and predicting survival outcomes, aiming to reduce organ discard rates and improve allocation efficiency beyond traditional scoring systems like the Kidney Donor Profile Index. Surgical data science utilizes AI to enhance robotic surgery through augmented reality for real-time anatomical visualization and 3D printed models for preoperative planning. Furthermore, ML is applied to assess organ quality during normothermic machine perfusion. Regarding post-transplant outcomes, artificial neural networks have demonstrated superior accuracy in predicting graft survival and rejection compared to conventional statistical methods. Despite these advancements, clinical application is hindered by limitations such as overfitting, selection bias from single-center data, and a lack of external validation.