Transductive image segmentation: Self-training and effect of uncertainty estimation

  title={Transductive image segmentation: Self-training and effect of uncertainty estimation},
  author={Konstantinos Kamnitsas and Stefan Winzeck and Evgenios N. Kornaropoulos and Daniel P Whitehouse and Cameron Englman and Poe ei Phyu and Norman Pao and David K. Menon and Daniel Rueckert and Tilak Das and Virginia F. J. Newcombe and Ben Glocker},
Semi-supervised learning (SSL) uses unlabeled data during training to learn better models. Previous studies on SSL for medical image segmentation focused mostly on improving model generalization to unseen data. In some applications, however, our primary interest is not generalization but to obtain optimal predictions on a specific unlabeled database that is fully available during model development. Examples include population studies for extracting imaging phenotypes. This work investigates an… 

Figures and Tables from this paper


Uncertainty-aware Self-ensembling Model for Semi-supervised 3D Left Atrium Segmentation
A novel uncertainty-aware semi-supervised framework for left atrium segmentation from 3D MR images that can effectively leverage the unlabeled data by encouraging consistent predictions of the same input under different perturbations.
Semi-supervised Learning for Network-Based Cardiac MR Image Segmentation
A semi-supervised learning approach, in which a segmentation network is trained from both labelled and unlabelled data, which outperforms a state-of-the-art multi-atlas segmentation method by a large margin and the speed is substantially faster.
Semi-Supervised Brain Lesion Segmentation with an Adapted Mean Teacher Model
This work proposes a semi-supervised learning (SSL) approach to brain lesion segmentation, where unannotated data is incorporated into the training of CNNs and outperforms competing SSL-based methods on ischemic stroke lesion segmentsation.
Omni-Supervised Learning: Scaling Up to Large Unlabelled Medical Datasets
This paper demonstrates the effectiveness of the proposed framework to localize multiple structures in a 3D US dataset of 4044 fetal brain volumes with an initial expert annotation of just 200 volumes, and seeks to exploit a small annotated dataset and iteratively increase model performance by scaling up to refine the model using a large set of unlabelled data.
Curriculum semi-supervised segmentation
This study investigates a curriculum-style strategy for semi-supervised CNN segmentation, which devises a regression network to learn image-level information such as the size of a target region, and achieves very competitive results, approaching the performance of full-supervision.
A New Ensemble Learning Framework for 3D Biomedical Image Segmentation
This paper develops a fully convolutional network based meta-learner to learn how to improve the results from 2D and 3D models, and devise a new training method that uses the results of the base-learners as multiple versions of "ground truths".
Progressive Graph-Based Transductive Learning for Multi-modal Classification of Brain Disorder Disease
A progressive GTL (pGTL) method that iteratively refines the subject-wise relationships observed in the feature domain using the learned intrinsic data representation in the label domain, and verifies the intrinsicData representation on the training data, in order to guarantee an optimal classification on the new testing data is proposed.
Realistic Evaluation of Deep Semi-Supervised Learning Algorithms
This work creates a unified reimplemention and evaluation platform of various widely-used SSL techniques and finds that the performance of simple baselines which do not use unlabeled data is often underreported, that SSL methods differ in sensitivity to the amount of labeled and unlabeling data, and that performance can degrade substantially when the unlabelED dataset contains out-of-class examples.
Analyzing the Quality and Challenges of Uncertainty Estimations for Brain Tumor Segmentation
An analysis of the most commonly used uncertainty estimation methods in regards to benefits and challenges for brain tumor segmentation, and shows that the uncertainty methods are typically well-calibrated when evaluated at the dataset level.
Semi-Supervised Learning
This first comprehensive overview of semi-supervised learning presents state-of-the-art algorithms, a taxonomy of the field, selected applications, benchmark experiments, and perspectives on ongoing and future research.