Abstract
In recent years, artificial intelligence (AI) has advanced significantly in the field of pathomics, enabling the transformation of pathological images into high-throughput, machine-readable datasets for quantitative analysis and precise diagnosis in renal pathology. AI-based pathomics has introduced innovative perspectives and sophisticated tools for detecting glomerular injury, evaluating renal interstitial fibrosis, monitoring transplanted kidney pathology longitudinally, and predicting outcomes in renal tumors. Despite its great promise, the application of AI-based pathomics in nephrology still faces several challenges, including complex data annotation, limited model interpretability, lack of comprehensive multi-modal data integration, and insufficient large-scale clinical validation. Future research should prioritize these challenges by enhancing multi-omics integration and promoting interdisciplinary collaboration, thereby advancing AI-based pathomics in nephrology and ultimately improving the precision and efficiency of patient care.