Bi-CRCL: Bidirectional Conservative-Radical Complementary Learning with Pre-trained Foundation Models for Class-incremental Medical Image Analysis

2026-03-24

AI Summary

This work addresses catastrophic forgetting in medical image class-incremental learning, a challenge exacerbated by data heterogeneity and privacy constraints that preclude data replay. To this end, the authors propose Bi-CRCL, a dual-learner framework that introduces the complementary learning systems paradigm into pre-trained foundation model–based incremental learning for medical imaging. The framework comprises a conservative learner that preserves stability to retain prior knowledge and a plasticity-enhanced radical learner that rapidly adapts to new classes. A bidirectional interaction mechanism facilitates forward transfer and backward consolidation between the two learners. Coupled with adaptive inference fusion and a replay-free training strategy, Bi-CRCL significantly outperforms existing methods across five medical imaging datasets, demonstrating robustness to cross-dataset distribution shifts and diverse task configurations while effectively balancing stability and plasticity.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Computer Vision: Medical and Biological Imaging
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Class-incremental learning (CIL) in medical image-guided diagnosis requires retaining prior diagnostic knowledge while adapting to newly emerging disease categories, which is critical for scalable clinical deployment. This problem is particularly challenging due to heterogeneous data and privacy constraints that prevent memory replay. Although pretrained foundation models (PFMs) have advanced general-domain CIL, their potential in medical imaging remains underexplored, where domain-specific adaptation is essential yet difficult due to anatomical complexity and inter-institutional heterogeneity. To address this gap, we conduct a systematic benchmark of recent PFM-based CIL methods and propose Bidirectional Conservative-Radical Complementary Learning (Bi-CRCL), a dual-learner framework inspired by complementary learning systems. Bi-CRCL integrates a conservative learner that preserves prior knowledge through stability-oriented updates and a radical learner that rapidly adapts to new categories via plasticity-oriented learning. A bidirectional interaction mechanism enables forward transfer and backward consolidation, allowing continual integration of new knowledge while mitigating catastrophic forgetting. During inference, outputs from both learners are adaptively fused for robust predictions. Experiments on five medical imaging datasets demonstrate consistent improvements over state-of-the-art methods under diverse settings, including cross-dataset shifts and varying task configurations.

Problem

• Class-incremental learning
• Medical image analysis
• Catastrophic forgetting
• Pre-trained foundation models
• Continual learning

Innovation

• Class-incremental learning
• Foundation models
• Complementary learning
• Catastrophic forgetting
• Medical image analysis