One-shot Weakly-Supervised Segmentation in Medical Images (2021-11-21T00:00:00.000000Z)

TL;DR

An innovative framework for 3D medical image segmentation with one-shot and weakly-supervised settings is presented and shows significant improvement over the state-of-the-art methods and performs robustly even under severe class imbalance and low contrast.

Abstract

Deep neural networks usually require accurate and a large number of annotations to achieve outstanding performance in medical image segmentation. One-shot segmentation and weakly-supervised learning are promising research directions that lower labeling effort by learning a new class from only one annotated image and utilizing coarse labels instead, respectively. Previous works usually fail to leverage the anatomical structure and suffer from class imbalance and low contrast problems. Hence, we present an innovative framework for 3D medical image segmentation with one-shot and weakly-supervised settings. Firstly a propagation-reconstruction network is proposed to project scribbles from annotated volume to unlabeled 3D images based on the assumption that anatomical patterns in different human bodies are similar. Then a dual-level feature denoising module is designed to refine the scribbles based on anatomical- and pixel-level features. After expanding the scribbles to pseudo masks, we could train a segmentation model for the new class with the noisy label training strategy. Experiments on one abdomen and one head-and-neck CT dataset show the proposed method obtains significant improvement over the state-of-the-art methods and performs robustly even under severe class imbalance and low contrast.

Authors

Guotai Wang

6 papers

Shaoting Zhang

7 papers

Wenhui Lei

2 papers

TL;DR

Abstract

Authors

References71 items

Recurrent Mask Refinement for Few-Shot Medical Image Segmentation

Distilling effective supervision for robust medical image segmentation with noisy labels

Hierarchical Consistency Regularized Mean Teacher for Semi-supervised 3D Left Atrium Segmentation

Adaptive Prototype Learning and Allocation for Few-Shot Segmentation

Learning From Pixel-Level Label Noise: A New Perspective for Semi-Supervised Semantic Segmentation

Learning with Feature-Dependent Label Noise: A Progressive Approach

One-Shot Medical Landmark Detection

Automatic Segmentation of Organs-at-Risk from Head-and-Neck CT using Separable Convolutional Neural Network with Hard-Region-Weighted Loss

Puzzle-CAM: Improved Localization Via Matching Partial And Full Features

Contrastive Learning of Relative Position Regression for One-Shot Object Localization in 3D Medical Images

SAM: Self-Supervised Learning of Pixel-Wise Anatomical Embeddings in Radiological Images

Self-Supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation

Contrastive learning of global and local features for medical image segmentation with limited annotations

A Noise-Robust Framework for Automatic Segmentation of COVID-19 Pneumonia Lesions From CT Images

Weakly Supervised Vessel Segmentation in X-ray Angiograms by Self-Paced Learning from Noisy Labels with Suggestive Annotation

Bounding boxes for weakly supervised segmentation: Global constraints get close to full supervision

Causality matters in medical imaging

Deep learning with noisy labels: exploring techniques and remedies in medical image analysis

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Interpreting chest X-rays via CNNs that exploit disease dependencies and uncertainty labels

DeepIGeoS-V2: Deep Interactive Segmentation of Multiple Organs from Head and Neck Images with Lightweight CNNs

Pancreatic Cancer Detection in Whole Slide Images Using Noisy Label Annotations

Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation

PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment

Uncertainty-aware Self-ensembling Model for Semi-supervised 3D Left Atrium Segmentation

Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction

Unequal-Training for Deep Face Recognition With Long-Tailed Noisy Data

Collaborative Learning of Semi-Supervised Segmentation and Classification for Medical Images

Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels

MixMatch: A Holistic Approach to Semi-Supervised Learning

Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation

Self-Supervised Visual Feature Learning With Deep Neural Networks: A Survey

'Squeeze & Excite' Guided Few-Shot Segmentation of Volumetric Images

Robust Inference via Generative Classifiers for Handling Noisy Labels

SG-One: Similarity Guidance Network for One-Shot Semantic Segmentation

Label Denoising with Large Ensembles of Heterogeneous Neural Networks

A Two-Stream Mutual Attention Network for Semi-Supervised Biomedical Segmentation with Noisy Labels

Weakly-supervised convolutional neural networks for multimodal image registration

Probabilistic Model-Agnostic Meta-Learning

Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing

Few-Shot Segmentation Propagation with Guided Networks

Task Agnostic Meta-Learning for Few-Shot Learning

Constrained‐CNN losses for weakly supervised segmentation☆

Co-teaching: Robust training of deep neural networks with extremely noisy labels

Training a neural network based on unreliable human annotation of medical images

Minimal annotation training for segmentation of microscopy images

Learning Pixel-Level Semantic Affinity with Image-Level Supervision for Weakly Supervised Semantic Segmentation

Learning to Reweight Examples for Robust Deep Learning

Learning to Compare: Relation Network for Few-Shot Learning

Automatic Brain Tumor Segmentation Using Cascaded Anisotropic Convolutional Neural Networks

One-Shot Learning for Semantic Segmentation

DeepIGeoS: A Deep Interactive Geodesic Framework for Medical Image Segmentation

Prototypical Networks for Few-shot Learning

Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

Meta Networks

Robust Loss Functions under Label Noise for Deep Neural Networks

One-Shot Video Object Segmentation

V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation

Matching Networks for One Shot Learning

DeepCut: Object Segmentation From Bounding Box Annotations Using Convolutional Neural Networks

One-shot Learning with Memory-Augmented Neural Networks

ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation

Simple Does It: Weakly Supervised Instance and Semantic Segmentation

Learning Deep Features for Discriminative Localization

U-Net: Convolutional Networks for Biomedical Image Segmentation

Adam: A Method for Stochastic Optimization

Learning from Noisy Labels with Deep Neural Networks

The Cancer Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository

GeoS: Geodesic Image Segmentation

Inter Extreme Points Geodesics for Weakly Supervised Segmentation

Uncertainty-aware self-ensembling model 11

Field of Study

Journal Information

Name

Volume

Venue Information