NORA: A Nephrology-Oriented Representation Learning Approach Towards Chronic Kidney Disease Classification

2025-09-16

AI Summary

Early detection of chronic kidney disease (CKD) in outpatient settings is hindered by limited access to kidney-specific biomarkers. To address this, we propose Nephrology-Oriented Representation leArning (NORA), a supervised contrastive learning framework that constructs disease-informed patient representations solely from routinely available, non-renal variables—including demographics, comorbidities, and urinalysis metrics—extracted from electronic health records. These representations are fed into a random forest classifier for automated CKD staging. Evaluated on the Riverside clinical dataset, NORA achieves a 12.3% improvement in F1-score for early CKD identification and demonstrates strong cross-institutional generalizability on the UCI CKD dataset. Crucially, NORA is the first method to explicitly embed domain knowledge into the contrastive learning objective, enabling high-accuracy, interpretable, and biomarker-free early CKD risk stratification—establishing a novel screening paradigm for resource-constrained healthcare settings.

Technology Category

• Knowledge Representation and Reasoning: Knowledge Acquisition
• Machine Learning: Representation Learning
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Chronic Kidney Disease (CKD) affects millions of people worldwide, yet its early detection remains challenging, especially in outpatient settings where laboratory-based renal biomarkers are often unavailable. In this work, we investigate the predictive potential of routinely collected non-renal clinical variables for CKD classification, including sociodemographic factors, comorbid conditions, and urinalysis findings. We introduce the Nephrology-Oriented Representation leArning (NORA) approach, which combines supervised contrastive learning with a nonlinear Random Forest classifier. NORA first derives discriminative patient representations from tabular EHR data, which are then used for downstream CKD classification. We evaluated NORA on a clinic-based EHR dataset from Riverside Nephrology Physicians. Our results demonstrated that NORA improves class separability and overall classification performance, particularly enhancing the F1-score for early-stage CKD. Additionally, we assessed the generalizability of NORA on the UCI CKD dataset, demonstrating its effectiveness for CKD risk stratification across distinct patient cohorts.

Problem

• Early CKD detection using non-renal clinical variables
• Improving classification with representation learning from EHR data
• Validating approach across different patient cohorts

Innovation

• Combines supervised contrastive learning
• Uses nonlinear Random Forest classifier
• Derives representations from tabular EHR