Unsupervised 3D End-to-End Medical Image Registration With Volume Tweening Network (2019-02-13T00:00:00.000000Z)

TL;DR

A new unsupervised learning method using convolutional neural networks under an end-to-end framework, Volume Tweening Network (VTN), for 3D medical image registration is proposed, which is 880x faster (or 3.3x faster without GPU acceleration) than traditional optimization-based methods and achieves state-of-the-art performance inmedical image registration.

Abstract

3D medical image registration is of great clinical importance. However, supervised learning methods require a large amount of accurately annotated corresponding control points (or morphing), which are very difficult to obtain. Unsupervised learning methods ease the burden of manual annotation by exploiting unlabeled data without supervision. In this article, we propose a new unsupervised learning method using convolutional neural networks under an end-to-end framework, Volume Tweening Network (VTN), for 3D medical image registration. We propose three innovative technical components: (1) An end-to-end cascading scheme that resolves large displacement; (2) An efficient integration of affine registration network; and (3) An additional invertibility loss that encourages backward consistency. Experiments demonstrate that our algorithm is 880x faster (or 3.3x faster without GPU acceleration) than traditional optimization-based methods and achieves state-of-the-art performance in medical image registration.

Authors

E. Chang

3 papers

Ting Fung Lau

1 papers

Ji Luo

1 papers

TL;DR

Abstract

Authors

References46 items

Advanced Normalization Tools (ANTs)

Recursive Cascaded Networks for Unsupervised Medical Image Registration

Networks for Joint Affine and Non-Parametric Image Registration

Learning a Probabilistic Model for Diffeomorphic Registration

A deep learning framework for unsupervised affine and deformable image registration

VoxelMorph: A Learning Framework for Deformable Medical Image Registration

Unsupervised Probabilistic Deformation Modeling for Robust Diffeomorphic Registration

An Unsupervised Learning Model for Deformable Medical Image Registration

Unsupervised End-to-end Learning for Deformable Medical Image Registration

Nonrigid Image Registration Using Multi-scale 3D Convolutional Neural Networks

End-to-End Unsupervised Deformable Image Registration with a Convolutional Neural Network

A survey on deep learning in medical image analysis

Unsupervised Deep Learning for Optical Flow Estimation

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

Unsupervised Monocular Depth Estimation with Left-Right Consistency

Back to Basics: Unsupervised Learning of Optical Flow via Brightness Constancy and Motion Smoothness

Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue

A CNN Regression Approach for Real-Time 2D/3D Registration

The Neuro Bureau ADHD-200 Preprocessed repository

Spatial Transformer Networks

U-Net: Convolutional Networks for Biomedical Image Segmentation

FlowNet: Learning Optical Flow with Convolutional Networks

Adam: A Method for Stochastic Optimization

Fully convolutional networks for semantic segmentation

Mutual Information between Discrete and Continuous Data Sets

Medical image registration: a review

Estimating the Unseen

Unsupervised Deep Feature Learning for Deformable Registration of MR Brain Images

The Autism Brain Imaging Data Exchange: Towards Large-Scale Evaluation of the Intrinsic Brain Architecture in Autism

Deformable Medical Image Registration: A Survey

Image similarity metrics in image registration

FAIR: Flexible Algorithms for Image Registration

Learning similarity measure for multi-modal 3D image registration

Comparison and Evaluation of Methods for Liver Segmentation From CT Datasets

Construction of a 3D probabilistic atlas of human cortical structures

Shape registration in implicit spaces using information theory and free form deformations

Ways toward an early diagnosis in Alzheimer’s disease: The Alzheimer’s Disease Neuroimaging Initiative (ADNI)

Mutual-information-based registration of medical images: a survey

Nonrigid registration using free-form deformations: application to breast MR images

Nonlinear total variation based noise removal algorithms

Liver tumor segmentation challenge

Vision – ECCV 2016

TensorFlow: Large-scale machine learning on heterogeneous systems

elastix: A Toolbox for Intensity-Based Medical Image Registration

c ○ 2002 Kluwer Academic Publishers. Manufactured in The Netherlands. Variational Methods for Multimodal Image Matching

An overlap invariant entropy measure of 3D medical image alignment

Field of Study

Journal Information

Name

Page

Volume

Venue Information

Name

Type

URL

Alternate Names