Abstract:Objective. In the field of medicine, semi-supervised segmentation algorithms hold crucial research significance while also facing substantial challenges, primarily due to the extreme scarcity of expert-level annotated medical image data. However, many existing semi-supervised methods still process labeled and unlabeled data in inconsistent ways, which can lead to knowledge learned from labeled data being discarded to some extent. This not only lacks a variety of perturbations to explore potential robust information in unlabeled data but also ignores the confirmation bias and class imbalance issues in pseudo-labeling methods. Approach. To solve these problems, this paper proposes a semi-supervised medical image segmentation method MDT that combines idea of consistency and pseudo-labeling. Firstly, MDT introduces a new perturbation strategy Mixup-Decoupling to fully regularize training data. It not only mixes labeled and unlabeled data at the data level but also performs decoupling operations between the output predictions of mixed target data and labeled data at the feature level to obtain strong version predictions of unlabeled data. Then it establishes a dual learning paradigm based on consistency and pseudo-labeling. Secondly, MDT employs a novel categorical entropy filtering approach to pick high-confidence pseudo-labels for unlabeled data, facilitating more refined supervision. Main results. This paper compares MDT with other advanced semi-supervised methods on 2D and 3D datasets separately. A large number of experimental results show that MDT achieves competitive segmentation performance and outperforms other state-of-the-art semi-supervised segmentation methods. Significance. This paper proposes a semi-supervised medical image segmentation method MDT, which greatly reduces the demand for manually labeled data and eases the difficulty of data annotation to a great extent. In addition, MDT not only outperforms many advanced semi-supervised image segmentation methods in quantitative and qualitative experimental results, but also provides a new and developable idea for semi-supervised learning and computer-aided diagnosis technology research.

Semi-supervised Meta-learning with Disentanglement for Domain-generalised Medical Image Segmentation

Semi-Supervised Convolutional Vision Transformer with Bi-Level Uncertainty Estimation for Medical Image Segmentation

Meta semi-supervised medical image segmentation with label hierarchy

SSL-DG: Rethinking and Fusing Semi-supervised Learning and Domain Generalization in Medical Image Segmentation

Unsupervised Domain Adaptation for Medical Image Segmentation by Selective Entropy Constraints and Adaptive Semantic Alignment

Meta-learners for few-shot weakly-supervised medical image segmentation

Domain Disentangled Meta-Learning

Disentangled Representation for Cross-Domain Medical Image Segmentation

Meta-Learned Feature Critics for Domain Generalized Semantic Segmentation

An Efficient Semi-Supervised Framework with Multi-Task and Curriculum Learning for Medical Image Segmentation

Transformation-Consistent Self-Ensembling Model for Semisupervised Medical Image Segmentation

SEMI-SUPERVISED DOMAIN GENERALIZATION FOR MEDICAL IMAGE ANALYSIS

LSRML: A latent space regularization based meta-learning framework for MR image segmentation

Consistency-guided Meta-Learning for Bootstrapping Semi-Supervised Medical Image Segmentation

Shape-aware Meta-learning for Generalizing Prostate MRI Segmentation to Unseen Domains

MDT: semi-supervised medical image segmentation with Mixup-Decoupling training

Transformation Consistent Self-ensembling Model for Semi-supervised Medical Image Segmentation

Exploring Generalizable Distillation for Efficient Medical Image Segmentation

Dual-consistency guidance semi-supervised medical image segmentation with low-level detail feature augmentation

Dual-Task Mutual Learning for Semi-Supervised Medical Image Segmentation