Evaluation and development of Artificial Intelligence tools to assess COVID-19 Severe Acute4 Respiratory Syndrome from chest imaging
Loading...
Date
Author(s)
Citation for Previous Publication
Link to Related Item
Abstract
Description
Chest CT is being more widely used as a diagnostic test for COVID-19 Severe Acute
Respiratory Syndrome-related lung disease. Artificial intelligence (AI) has the ability
to assist in the rapid assessment of CT scans for COVID-19 Severe Acute Respiratory
Syndrome findings from other clinical entities. Here that a group of deep learning al-
gorithms trained on a diverse multinational cohort of 1280 patients to localize parietal
lung parenchyma followed by classification of COVID-19 Severe Acute Respiratory
Syndrome can achieve up to 90.8 percent accuracy, with 84 percent sensitivity and 93
percent specificity, as measured in an independent test set (not included in training
and validation) can achieve up to 90.8 percent accuracy, with 84 percent sensitivity
and 93 percent specificity. Chest CTs from oncology, emergency, and pneumonia-
related indications were used as normal controls. In 140 patients with laboratory
reported other (non COVID-19) pneumonia, the false positive rate was 10 percent.
In a variety of patient populations, AI-based algorithms can quickly differentiate CT
scans with COVID-19 induced pneumonia, as well as distinguish non-COVID related
Severe Acute Respiratory Syndrome from chest imaging with high specificity.
Item Type
http://purl.org/coar/resource_type/c_1843
Alternative
Other License Text / Link
Language
en