SVF-Net: Learning Deformable Image Registration Using Shape Matching

Abstract

Abstract. In this paper, we propose an innovative approach for registration based on the deterministic prediction of the parameters from both images instead of the optimization of a energy criteria. The method relies on a fully convolutional network whose architecture consists of contracting layers to detect relevant features and a symmetric expanding path that matches them together and outputs the transformation parametrization. Whereas convolutional networks have seen a widespread expansion and have been already applied to many medical imaging problems such as segmentation and classification, its application to registration has so far faced the challenge of defining ground truth data on which to train the algorithm. Here, we present a novel training strategy to build reference deformations which relies on the registration of segmented regions of interest. We apply this methodology to the problem of inter-patient heart registration and show an important improvement over a state of the art optimization based algorithm. Not only our method is more accurate but it is also faster -registration of two 3D-images taking less than 30ms second on a GPU -and more robust to outliers.

Learning deformable registration of medical images with anatomical constraints

“…In particular, deep convolutional neural networks (CNN) have proved to outperform all existent strategies in other fundamental tasks of computer vision, like image segmentation [27] and classification [23]. During the last years, we have witnessed the advent of deep learning-based image registration methods [25,49,35,39,46,45,1,6,11], which achieve state-of-the-art performance, and drastically reduce the required computational time. These works have made a fundamental contribution by setting novel architectures for CNN-based deformable image registration (following supervised, unsupervised and semi-supervised training approaches).…”

Fast contour propagation for MR‐guided prostate radiotherapy using convolutional neural networks

“…Unsupervised methods learn the deformation directly from pairs of images without a ground truth deformation vector field by maximizing a similarity metric . Strongly supervised methods use a ground truth deformation vector field, usually by applying known transformations to a set of images during training . Weakly supervised methods are a variant of unsupervised methods, in which the similarity metric is replaced by learning an auxiliary task, such as maximizing the overlap of known segmentations .…”

Probabilistic Multilayer Regularization Network for Unsupervised 3D Brain Image Registration

“…For example, Sokooti et al [16] developed a RegNet trained with the generated displacement vector fields to register CT images. Rohé et al [14] learned to align the images by leveraging additional shape priors in a CNN. However, these methods leverage the manually-labeled ground truth to train the deep networks in a supervised manner, where the labeled images are expensive and tedious to be obtained.…”