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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Good quality of medical images is a prerequisite for the success ofsubsequent image analysis pipelines. Quality assessment of medical images istherefore an essential activity and for large population studies such as the UKBiobank (UKBB), manual identification of artefacts such as those caused byunanticipated motion is tedious and time-consuming. Therefore, there is anurgent need for automatic image quality assessment techniques. In this paper,we propose a method to automatically detect the presence of motion-relatedartefacts in cardiac magnetic resonance (CMR) cine images. We compare two deeplearning architectures to classify poor quality CMR images: 1) 3Dspatio-temporal Convolutional Neural Networks (3D-CNN), 2) Long-term RecurrentConvolutional Network (LRCN). Though in real clinical setup motion artefactsare common, high-quality imaging of UKBB, which comprises cross-sectionalpopulation data of volunteers who do not necessarily have health problemscreates a highly imbalanced classification problem. Due to the high number ofgood quality images compared to the relatively low number of images with motionartefacts, we propose a novel data augmentation scheme based on syntheticartefact creation in k-space. We also investigate a learning approach using apredetermined curriculum based on synthetic artefact severity. We evaluate ourpipeline on a subset of the UK Biobank data set consisting of 3510 CMR images.The LRCN architecture outperformed the 3D-CNN architecture and was able todetect 2D+time short axis images with motion artefacts in less than 1ms withhigh recall. We compare our approach to a range of state-of-the-art qualityassessment methods. The novel data augmentation and curriculum learningapproaches both improved classification performance achieving overall areaunder the ROC curve of 0.89.

Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning