Continual Active Learning for Efficient Adaptation of Machine Learning Models to Changing Image Acquisition

TL;DR

This paper introduces a continual active learning method for medical imaging that adapts to changing data sources and reduces manual labeling, improving model accuracy in dynamic clinical environments.

Contribution

It presents a novel continual active learning approach that detects domain shifts, adapts training, and efficiently selects samples for labeling in medical imaging scenarios.

Findings

Outperforms naive active learning methods.

Requires less manual labeling.

Effective in adapting to changing imaging sources.

Abstract

Imaging in clinical routine is subject to changing scanner protocols, hardware, or policies in a typically heterogeneous set of acquisition hardware. Accuracy and reliability of deep learning models suffer from those changes as data and targets become inconsistent with their initial static training set. Continual learning can adapt to a continuous data stream of a changing imaging environment. Here, we propose a method for continual active learning on a data stream of medical images. It recognizes shifts or additions of new imaging sources - domains -, adapts training accordingly, and selects optimal examples for labelling. Model training has to cope with a limited labelling budget, resembling typical real world scenarios. We demonstrate our method on T1-weighted magnetic resonance images from three different scanners with the task of brain age estimation. Results demonstrate that the proposed method outperforms naive active learning while requiring less manual labelling.