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

The human lung is a complex respiratory organ, consisting of five distinctanatomic compartments called lobes. Accurate and automatic segmentation ofthese pulmonary lobes from computed tomography (CT) images is of clinicalimportance for lung disease assessment and treatment planning. However, thistask is challenging due to ambiguous lobar boundaries, anatomical variationsand pathological deformations. In this paper, we propose a high-resolution andefficient 3D fully convolutional network to automatically segment the lobes. Werefer to the network as Pulmonary Lobe Segmentation Network (PLS-Net), which isdesigned to efficiently exploit 3D spatial and contextual information fromhigh-resolution volumetric CT images for effective volume-to-volume learningand inference. The PLS-Net is based on an asymmetric encoder-decoderarchitecture with three novel components: (i) 3D depthwise separableconvolutions to improve the network efficiency by factorising each regular 3Dconvolution into two simpler operations; (ii) dilated residual dense blocks toefficiently expand the receptive field of the network and aggregate multi-scalecontextual information for segmentation; and (iii) input reinforcement at eachdownsampled resolution to compensate for the loss of spatial information due toconvolutional and downsampling operations. We evaluated the proposed PLS-Net ona multi-institutional dataset that consists of 210 CT images acquired frompatients with a wide range of lung abnormalities. Experimental results showthat our PLS-Net achieves state-of-the-art performance with bettercomputational efficiency. Further experiments confirm the effectiveness of eachnovel component of the PLS-Net.

Efficient 3D Fully Convolutional Networks for Pulmonary Lobe Segmentation in CT Images