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

With the advent of Cardiac Cine Magnetic Resonance (CMR) Imaging, there hasbeen a paradigm shift in medical technology, thanks to its capability ofimaging different structures within the heart without ionizing radiation.However, it is very challenging to conduct pre-operative planning of minimallyinvasive cardiac procedures without accurate segmentation and identification ofthe left ventricle (LV), right ventricle (RV) blood-pool, and LV-myocardium.Manual segmentation of those structures, nevertheless, is time-consuming andoften prone to error and biased outcomes. Hence, automatic and computationallyefficient segmentation techniques are paramount. In this work, we propose anovel memory-efficient Convolutional Neural Network (CNN) architecture as amodification of both CondenseNet, as well as DenseNet for ventricularblood-pool segmentation by introducing a bottleneck block and an upsamplingpath. Our experiments show that the proposed architecture runs on the AutomatedCardiac Diagnosis Challenge (ACDC) dataset using half (50%) the memoryrequirement of DenseNet and one-twelfth (~ 8%) of the memory requirements ofU-Net, while still maintaining excellent accuracy of cardiac segmentation. Wevalidated the framework on the ACDC dataset featuring one healthy and fourpathology groups whose heart images were acquired throughout the cardiac cycleand achieved the mean dice scores of 96.78% (LV blood-pool), 93.46% (RVblood-pool) and 90.1% (LV-Myocardium). These results are promising and promotethe proposed methods as a competitive tool for cardiac image segmentation andclinical parameter estimation that has the potential to provide fast andaccurate results, as needed for pre-procedural planning and/or pre-operativeapplications.

CondenseUNet: A Memory-Efficient Condensely-Connected Architecture for Bi-ventricular Blood Pool and Myocardium Segmentation