PAGE: Domain-Incremental Adaptation with Past-Agnostic Generative Replay for Smart Healthcare
Chia‐Hao Li, Niraj Kumar Jha · ACM Transactions on Computing for Healthcare · 2025
Modern advances in machine learning (ML) and wearable medical sensors (WMSs) have enabled out-of-clinic disease detection. However, a trained ML model often suffers from misclassification when encountering non-stationary data domains after deployment. Use of continual learning (CL) strategies is a common way to perform domain-incremental adaptation while mitigating catastrophic forgetting. Nevertheless, most existing CL methods require access to previously learned domains through preservation of raw training data or distilled information. This is often infeasible in real-world scenarios due to storage limitations or data privacy, especially in smart healthcare applications. Moreover, it makes most existing CL algorithms inapplicable to deployed models in the field, thus incurring re-engineering costs. To address these challenges, we propose PAGE, a domain-incremental adaptation strategy with past-agnostic generative replay for smart healthcare. PAGE enables generative replay without the aid of any preserved data or information from prior domains. When adapting to a new domain, it exploits real data from the new distribution and the current model to generate synthetic data that retain the learned knowledge of previous domains. By replaying the synthetic data with the new real data during training, PAGE achieves a good balance between domain adaptation and knowledge retention. In addition, we incorporate an extended inductive conformal prediction (EICP) method into PAGE to produce a confidence score and a credibility value for each detection result. This makes the predictions interpretable and provides statistical guarantees for disease detection in smart healthcare applications. We demonstrate PAGE’s effectiveness in domain-incremental disease detection with three distinct disease datasets collected from commercially available WMSs. PAGE achieves highly competitive performance against state-of-the-art along with superior scalability, data privacy, and feasibility. Furthermore, PAGE is able to enable up to 75% reduction in clinical workload with the help of EICP.