Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2021 Oct 14;11:20384. doi: 10.1038/s41598-021-99986-3

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

© The Author(s) 2021

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

PMC Copyright notice

Constituents of the CovIx ensemble. The low- and high-resolution image-wise classifiers were trained on frontal CXRs scaled to 299 × 299 and 764 × 764 pixels respectively. The classification head (H) contained two outputs —an NLP multi-label classifier output (L1-LN) and a COVID-19 classifier (Softmax). The NLP output consisted of a Dense layer with a neuron per NLP target class (classes = 10) followed by a Sigmoid activation function, while the COVID-19 classifier output likewise consisted of a Dense layer with four output neurons representing Normal, Abnormal, Pneumonia and COVID + respectively followed by a Softmax output. The patch-wise classifier was built by scaling each image to 1500 × 1500 resolution, extracting lung and heart masks, and taking 50 random patches cropped to image masks with a size of 299 × 299 as the network inputs. At inference stage, 50 random patches were acquired for each image and fed to the classifier to generate class probability values for Normal, Abnormal, Pneumonia, and COVID + classes.