Using Predictive Models and AI for AKI Research

Abstract

Acute kidney injury (AKI), a drop in kidney function with multiple etiologies, is a common complication in hospitalized patients and is associated with poorer patient outcomes. With the advent of electronic health records, machine learning algorithms have been developed that can predict the incidence and severity of AKI, AKI persistence, as well as patient outcomes like mortality and the need for kidney replacement therapies. Furthermore, it can risk-stratify patients based on early presentations to aid with clinical management. Newer technologies like natural language processing and generative artificial intelligence (AI) (e.g., ChatGPT) also show promise in the realm of AKI prediction and management. This review provides an overview of the role of AI in adults with AKI, as well as explores some limitations and ethical considerations that need to be addressed as we move forward.