English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Diffeomorphic deformable image registration is one of the crucial tasks inmedical image analysis, which aims to find a unique transformation whilepreserving the topology and invertibility of the transformation. Deepconvolutional neural networks (CNNs) have yielded well-suited approaches forimage registration by learning the transformation priors from a large dataset.The improvement in the performance of these methods is related to their abilityto learn information from several sample medical images that are difficult toobtain and bias the framework to the specific domain of data. In this paper, wepropose a novel diffeomorphic training-free approach; this is built upon theprinciple of an ordinary differential equation.  Our formulation yields an Euler integration type recursive scheme to estimatethe changes of spatial transformations between the fixed and the moving imagepyramids at different resolutions. The proposed architecture is simple indesign. The moving image is warped successively at each resolution and finallyaligned to the fixed image; this procedure is recursive in a way that at eachresolution, a fully convolutional network (FCN) models a progressive change ofdeformation for the current warped image. The entire system is end-to-end andoptimized for each pair of images from scratch. In comparison to learning-basedmethods, the proposed method neither requires a dedicated training set norsuffers from any training bias. We evaluate our method on three cardiac imagedatasets. The evaluation results demonstrate that the proposed method achievesstate-of-the-art registration accuracy while maintaining desirablediffeomorphic properties.

A training-free recursive multiresolution framework for diffeomorphic deformable image registration