Semi-Supervised Medical Image Segmentation via Learning Consistency under Transformations

Abstract

The scarcity of labeled data often limits the application of supervised deep learning techniques for medical image segmentation. This has motivated the development of semi-supervised techniques that learn from a mixture of labeled and unlabeled images. In this paper, we propose a novel semi-supervised method that, in addition to supervised learning on labeled training images, learns to predict segmentations consistent under a given class of transformations on both labeled and unlabeled images. More specifically, in this work we explore learning equivariance to elastic deformations. We implement this through: 1) a Siamese architecture with two identical branches, each of which receives a differently transformed image, and 2) a composite loss function with a supervised segmentation loss term and an unsupervised term that encourages segmentation consistency between the predictions of the two branches. We evaluate the method on a public dataset of chest radiographs with segmentations of anatomical structures using 5-fold cross-validation. The proposed method reaches significantly higher segmentation accuracy compared to supervised learning. This is due to learning transformation consistency on both labeled and unlabeled images, with the latter contributing the most. We achieve the performance comparable to state-of-the-art chest X-ray segmentation methods while using substantially fewer labeled images.

References in corpus (6)

• Semi-Supervised Learning with Ladder Networks
• Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning
• Semi-Supervised Deep Learning for Fully Convolutional Networks
• Improving the Segmentation of Anatomical Structures in Chest Radiographs using U-Net with an ImageNet Pre-trained Encoder
• Deep semi-supervised segmentation with weight-averaged consistency targets
• Semi-supervised Skin Lesion Segmentation via Transformation Consistent Self-ensembling Model

Cited by in corpus (25)

• Deep Learning for Chest X-ray Analysis: A Survey
• Shape-aware Semi-supervised 3D Semantic Segmentation for Medical Images
• Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation
• Self-paced and self-consistent co-training for semi-supervised image segmentation
• Active, Continual Fine Tuning of Convolutional Neural Networks for Reducing Annotation Efforts
• Which Pixel to Annotate: a Label-Efficient Nuclei Segmentation Framework
• Exploring Feature Representation Learning for Semi-supervised Medical Image Segmentation
• Learning Inductive Attention Guidance for Partially Supervised Pancreatic Ductal Adenocarcinoma Prediction
• DS6, Deformation-aware Semi-supervised Learning: Application to Small Vessel Segmentation with Noisy Training Data
• Efficient Semi-Supervised Gross Target Volume of Nasopharyngeal Carcinoma Segmentation via Uncertainty Rectified Pyramid Consistency
• Longitudinal diffusion MRI analysis using Segis-Net: a single-step deep-learning framework for simultaneous segmentation and registration
• Boosting Semi-supervised Image Segmentation with Global and Local Mutual Information Regularization
• Multi-organ segmentation: a progressive exploration of learning paradigms under scarce annotation
• Semi-supervised Active Learning for Instance Segmentation via Scoring Predictions
• Self-Paced Contrastive Learning for Semi-supervised Medical Image Segmentation with Meta-labels
• Supervision by Denoising for Medical Image Segmentation
• Contrastive Semi-Supervised Learning for 2D Medical Image Segmentation
• Uncertainty-Aware Self-supervised Neural Network for Liver $T_{1\rho }$ Mapping with Relaxation Constraint
• Semi-Supervised Medical Image Segmentation with Co-Distribution Alignment
• Revisiting consistency for semi-supervised semantic segmentation
• Towards Label-Free 3D Segmentation of Optical Coherence Tomography Images of the Optic Nerve Head Using Deep Learning
• ROAM: Random Layer Mixup for Semi-Supervised Learning in Medical Imaging
• Diversified Multi-prototype Representation for Semi-supervised Segmentation
• Medical Image Segmentation with Limited Supervision: A Review of Deep Network Models
• Context-aware virtual adversarial training for anatomically-plausible segmentation