Scribble-based Domain Adaptation via Co-segmentation
This addresses the challenge of domain adaptation in medical imaging for clinicians and researchers, reducing annotation effort while maintaining accuracy, though it is incremental as it builds on existing weakly-supervised and co-segmentation ideas.
The paper tackles the problem of poor generalization in medical image segmentation across different imaging modalities by proposing a weakly-supervised domain adaptation method using scribbles instead of full annotations. It outperforms unsupervised approaches and achieves performance comparable to fully-supervised methods on Vestibular Schwannoma segmentation from T1 to T2 scans.
Although deep convolutional networks have reached state-of-the-art performance in many medical image segmentation tasks, they have typically demonstrated poor generalisation capability. To be able to generalise from one domain (e.g. one imaging modality) to another, domain adaptation has to be performed. While supervised methods may lead to good performance, they require to fully annotate additional data which may not be an option in practice. In contrast, unsupervised methods don't need additional annotations but are usually unstable and hard to train. In this work, we propose a novel weakly-supervised method. Instead of requiring detailed but time-consuming annotations, scribbles on the target domain are used to perform domain adaptation. This paper introduces a new formulation of domain adaptation based on structured learning and co-segmentation. Our method is easy to train, thanks to the introduction of a regularised loss. The framework is validated on Vestibular Schwannoma segmentation (T1 to T2 scans). Our proposed method outperforms unsupervised approaches and achieves comparable performance to a fully-supervised approach.