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 current COVID-19 pandemic overloads healthcare systems, includingradiology departments. Though several deep learning approaches were developedto assist in CT analysis, nobody considered study triage directly as a computerscience problem. We describe two basic setups: Identification of COVID-19 toprioritize studies of potentially infected patients to isolate them as early aspossible; Severity quantification to highlight studies of severe patients anddirect them to a hospital or provide emergency medical care. We formalize thesetasks as binary classification and estimation of affected lung percentage.Though similar problems were well-studied separately, we show that existingmethods provide reasonable quality only for one of these setups. We employ amultitask approach to consolidate both triage approaches and propose aconvolutional neural network to combine all available labels within a singlemodel. In contrast with the most popular multitask approaches, we addclassification layers to the most spatially detailed upper part of U-Netinstead of the bottom, less detailed latent representation. We train our modelon approximately 2000 publicly available CT studies and test it with acarefully designed set consisting of 32 COVID-19 studies, 30 cases withbacterial pneumonia, 31 healthy patients, and 30 patients with other lungpathologies to emulate a typical patient flow in an out-patient hospital. Theproposed multitask model outperforms the latent-based one and achieves ROC AUCscores ranging from 0.87+-01 (bacterial pneumonia) to 0.97+-01 (healthycontrols) for Identification of COVID-19 and 0.97+-01 Spearman Correlation forSeverity quantification. We release all the code and create a publicleaderboard, where other community members can test their models on our testdataset.

CT-based COVID-19 Triage: Deep Multitask Learning Improves Joint Identification and Severity Quantification