Task-Agnostic Continual Learning for Chest Radiograph Classification
2026-02-17 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors look at how to update chest X-ray diagnostic models when new data comes in, without needing to retrain from scratch or lose accuracy. They introduce a method called CARL-XRay, which keeps a fixed main model and adds small task-specific parts for each new dataset. A smart selector helps the model figure out which task it is working on without needing to store all past images. Their experiments show CARL-XRay works better than training all tasks together, especially when the model doesn't know the task in advance, while using fewer resources. This approach could help deploy medical imaging AI that adapts over time more practically.
chest radiograph classificationcontinual learningtask-incremental learningadapter-based learningtask selectorfeature-level experience replayAUROClatent task identificationjoint training
Authors
Muthu Subash Kavitha, Anas Zafar, Amgad Muneer, Jia Wu
Abstract
Clinical deployment of chest radiograph classifiers requires models that can be updated as new datasets become available without retraining on previously ob- served data or degrading validated performance. We study, for the first time, a task-incremental continual learning setting for chest radiograph classification, in which heterogeneous chest X-ray datasets arrive sequentially and task identifiers are unavailable at inference. We propose a continual adapter-based routing learning strategy for Chest X-rays (CARL-XRay) that maintains a fixed high-capacity backbone and incrementally allocates lightweight task-specific adapters and classifier heads. A latent task selector operates on task-adapted features and leverages both current and historical context preserved through compact prototypes and feature-level experience replay. This design supports stable task identification and adaptation across sequential updates while avoiding raw-image storage. Experiments on large-scale public chest radiograph datasets demonstrate robust performance retention and reliable task-aware inference under continual dataset ingestion. CARL-XRay outperforms joint training under task-unknown deployment, achieving higher routing accuracy (75.0\% vs.\ 62.5\%), while maintaining competitive diagnostic performance with AUROC of 0.74 in the oracle setting with ground-truth task identity and 0.75 under task-unknown inference, using significantly fewer trainable parameters. Finally, the proposed framework provides a practical alternative to joint training and repeated full retraining in continual clinical deployment.