Publication:

Artificial intelligence approaches in early detection and clinical management of acute and chronic kidney diseases

Acute kidney injury (AKI) and chronic kidney disease (CKD) are major global health burdens that are often detected late due to the limitations of conventional biomarkers such as serum creatinine and estimated glomerular filtration rate. Artificial intelligence (AI) has emerged as a powerful tool capable of analyzing complex, high-dimensional clinical data to improve risk stratification, early detection, prognosis and personalized management of kidney diseases. This review evaluates the current applications of AI in the diagnosis, prediction and clinical management of AKI and CKD and compares its performance with traditional diagnostic approaches. A comprehensive literature search was conducted up to October 2025 using Google Scholar, PubMed, Web of Science and Scopus. Studies focusing on AI-based predictive modeling, imaging analysis, biomarker discovery and clinical decision support in nephrology were included. Of 150 screened articles, 51 met the inclusion criteria.AI-based models demonstrated superior accuracy for early AKI detection compared with serum creatinine alone (AUC>0.85 vs. 0.65) and improved prediction of CKD progression, cardiovascular outcomes, and dialysis initiation. Additionally, AI-assisted imaging enhanced renal pathology detection, while decision-support systems optimized drug dosing and dialysis parameters. Despite these advances, challenges such as algorithmic bias, interpretability, data heterogeneity, and ethical concerns remain. Overall, AI holds substantial potential to transform nephrology through earlier diagnosis, improved prognostication, and individualized treatment, though robust regulatory frameworks and large prospective studies are essential before widespread clinical implementation.