Leveraging Old Knowledge to Continually Learn New Classes in Medical Images

TL;DR

This paper presents a continual learning framework for medical image classification that uses a dynamic architecture and a balanced training procedure to learn new diseases without forgetting old ones, outperforming existing methods.

Contribution

It introduces a novel framework combining expanding representations and an alternating training strategy for effective class-incremental learning in medical imaging.

Findings

Achieves higher class accuracy than baselines.

Reduces catastrophic forgetting effectively.

Demonstrates robustness across multiple medical datasets.

Abstract

Class-incremental continual learning is a core step towards developing artificial intelligence systems that can continuously adapt to changes in the environment by learning new concepts without forgetting those previously learned. This is especially needed in the medical domain where continually learning from new incoming data is required to classify an expanded set of diseases. In this work, we focus on how old knowledge can be leveraged to learn new classes without catastrophic forgetting. We propose a framework that comprises of two main components: (1) a dynamic architecture with expanding representations to preserve previously learned features and accommodate new features; and (2) a training procedure alternating between two objectives to balance the learning of new features while maintaining the model's performance on old classes. Experiment results on multiple medical datasets show that our solution is able to achieve superior performance over state-of-the-art baselines in terms of class accuracy and forgetting.