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

Recently there has been an explosion in the use of Deep Learning (DL) methodsfor medical image segmentation. However the field's reliability is hindered bythe lack of a common base of reference for accuracy/performance evaluation andthe fact that previous research uses different datasets for evaluation. In thispaper, an extensive comparison of DL models for lung and COVID-19 lesionsegmentation in Computerized Tomography (CT) scans is presented, which can alsobe used as a benchmark for testing medical image segmentation models. Four DLarchitectures (Unet, Linknet, FPN, PSPNet) are combined with 25 randomlyinitialized and pretrained encoders (variations of VGG, DenseNet, ResNet,ResNext, DPN, MobileNet, Xception, Inception-v4, EfficientNet), to construct200 tested models. Three experimental setups are conducted for lungsegmentation, lesion segmentation and lesion segmentation using the originallung masks. A public COVID-19 dataset with 100 CT scan images (80 for train, 20for validation) is used for training/validation and a different public datasetconsisting of 829 images from 9 CT scan volumes for testing. Multiple findingsare provided including the best architecture-encoder models for each experimentas well as mean Dice results for each experiment, architecture and encoderindependently. Finally, the upper bounds improvements when using lung masks asa preprocessing step or when using pretrained models are quantified. The sourcecode and 600 pretrained models for the three experiments are provided, suitablefor fine-tuning in experimental setups without GPU capabilities.

Comprehensive Comparison of Deep Learning Models for Lung and COVID-19 Lesion Segmentation in CT scans