Chronic Kidney Disease Prognosis Prediction Using Transformer

TL;DR

This paper introduces ProQ-BERT, a transformer-based model that predicts CKD progression using multi-modal EHR data, achieving high accuracy and interpretability for personalized prognosis.

Contribution

The study develops a novel transformer framework with quantization tokenization and attention mechanisms for CKD prognosis prediction from multi-modal EHR data.

Findings

ProQ-BERT outperforms CEHR-BERT with ROC-AUC up to 0.995.

Achieved PR-AUC up to 0.989 for short-term CKD progression.

Effective use of transformer architecture for clinical prognosis modeling.

Abstract

Chronic Kidney Disease (CKD) affects nearly 10\% of the global population and often progresses to end-stage renal failure. Accurate prognosis prediction is vital for timely interventions and resource optimization. We present a transformer-based framework for predicting CKD progression using multi-modal electronic health records (EHR) from the Seoul National University Hospital OMOP Common Data Model. Our approach (\textbf{ProQ-BERT}) integrates demographic, clinical, and laboratory data, employing quantization-based tokenization for continuous lab values and attention mechanisms for interpretability. The model was pretrained with masked language modeling and fine-tuned for binary classification tasks predicting progression from stage 3a to stage 5 across varying follow-up and assessment periods. Evaluated on a cohort of 91,816 patients, our model consistently outperformed CEHR-BERT, achieving ROC-AUC up to 0.995 and PR-AUC up to 0.989 for short-term prediction. These results highlight the effectiveness of transformer architectures and temporal design choices in clinical prognosis modeling, offering a promising direction for personalized CKD care.