Medical Image Segmentation with Limited Supervision: A Review of Deep Network Models

Abstract

Despite the remarkable performance of deep learning methods on various tasks, most cutting-edge models rely heavily on large-scale annotated training examples, which are often unavailable for clinical and health care tasks. The labeling costs for medical images are very high, especially in medical image segmentation, which typically requires intensive pixel/voxel-wise labeling. Therefore, the strong capability of learning and generalizing from limited supervision, including a limited amount of annotations, sparse annotations, and inaccurate annotations, is crucial for the successful application of deep learning models in medical image segmentation. However, due to its intrinsic difficulty, segmentation with limited supervision is challenging and specific model design and/or learning strategies are needed. In this paper, we provide a systematic and up-to-date review of the solutions above, with summaries and comments about the methodologies. We also highlight several problems in this field, discussed future directions observing further investigations.

References in corpus (61)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• A Survey on Deep Learning in Medical Image Analysis
• Conditional Generative Adversarial Nets
• mixup: Beyond Empirical Risk Minimization
• Attention U-Net: Learning Where to Look for the Pancreas
• How transferable are features in deep neural networks?
• Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• Improved Regularization of Convolutional Neural Networks with Cutout
• Generative Adversarial Network in Medical Imaging: A Review
• Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
• Temporal Ensembling for Semi-Supervised Learning
• Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: the LUNA16 challenge
• Weight Uncertainty in Neural Networks
• The Liver Tumor Segmentation Benchmark (LiTS)
• Interactive Medical Image Segmentation using Deep Learning with Image-specific Fine-tuning
• A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration
• Learning to Reweight Examples for Robust Deep Learning
• Deep neural network models for computational histopathology: A survey
• Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentation
• Random Erasing Data Augmentation
• Meta-Learning for Semi-Supervised Few-Shot Classification
• Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks
• Transfusion: Understanding Transfer Learning for Medical Imaging
• Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes
• Deep Label Distribution Learning with Label Ambiguity
• Automatic Brain Tumor Segmentation using Cascaded Anisotropic Convolutional Neural Networks
• DeepIGeoS: A Deep Interactive Geodesic Framework for Medical Image Segmentation
• Semantic Segmentation using Adversarial Networks
• Manifold Mixup: Better Representations by Interpolating Hidden States
• Big Self-Supervised Models are Strong Semi-Supervised Learners
• Adversarial Attacks and Defences: A Survey
• Shape-aware Semi-supervised 3D Semantic Segmentation for Medical Images
• MILD-Net: Minimal Information Loss Dilated Network for Gland Instance Segmentation in Colon Histology Images
• Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation
• The KiTS19 Challenge Data: 300 Kidney Tumor Cases with Clinical Context, CT Semantic Segmentations, and Surgical Outcomes
• Patch-based Output Space Adversarial Learning for Joint Optic Disc and Cup Segmentation
• Constrained-CNN losses for weakly supervised segmentation
• Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptation
• A Topological Loss Function for Deep-Learning based Image Segmentation using Persistent Homology
• Semi-Supervised Medical Image Segmentation via Learning Consistency under Transformations
• On The Power of Curriculum Learning in Training Deep Networks
• Semi and Weakly Supervised Semantic Segmentation Using Generative Adversarial Network
• Prior-aware Neural Network for Partially-Supervised Multi-Organ Segmentation
• Weakly Supervised Vessel Segmentation in X-ray Angiograms by Self-Paced Learning from Noisy Labels with Suggestive Annotation
• Adversarial-Prediction Guided Multi-task Adaptation for Semantic Segmentation of Electron Microscopy Images
• A Survey on Deep Learning of Small Sample in Biomedical Image Analysis
• Uncertainty-Guided Efficient Interactive Refinement of Fetal Brain Segmentation from Stacks of MRI Slices
• Unsupervised Domain Adaptation in Semantic Segmentation: a Review
• Bounding boxes for weakly supervised segmentation: Global constraints get close to full supervision
• A sparse annotation strategy based on attention-guided active learning for 3D medical image segmentation
• Self-Loop Uncertainty: A Novel Pseudo-Label for Semi-Supervised Medical Image Segmentation
• Learning from Multiple Datasets with Heterogeneous and Partial Labels for Universal Lesion Detection in CT
• DoDNet: Learning to segment multi-organ and tumors from multiple partially labeled datasets
• Convex Shape Prior for Deep Neural Convolution Network based Eye Fundus Images Segmentation
• Deep Convolutional Neural Networks with Spatial Regularization, Volume and Star-shape Priori for Image Segmentation
• Marginal loss and exclusion loss for partially supervised multi-organ segmentation
• Extreme Consistency: Overcoming Annotation Scarcity and Domain Shifts
• Imperfect Segmentation Labels: How Much Do They Matter?
• Deep Learning from Small Amount of Medical Data with Noisy Labels: A Meta-Learning Approach