What You Should Know:
– New RADLogics research that validates the performance of an AI-powered CT image analysis solution that is designed to automatically and accurately detect COVID-19 (Coronavirus) and quantify the disease burden in affected patients.
– The study found that the CT image analysis algorithm – developed from multiple international datasets – was able to differentiate 157 patients with and without COVID-19 with a 0.996 AUC (plus, 98.2 percent sensitivity and 92.2 percent specificity).
– Although it is not recommended as a first-line test, non-contrast thoracic CT has been shown to be an effective tool in the detection, quantification, and follow-up of COVID-19.
RADLogics announced today new research that validates the performance of an AI-powered CT image analysis solution that is designed to automatically and accurately detect COVID-19 (Coronavirus) and quantify the disease burden in affected patients. To meet the growing worldwide pandemic, RADLogics also announced that it has rapidly deployed this new CT image analysis algorithm, which helps classify results for patients with COVID-19 per thoracic CT studies utilizing deep-learning image analysis.
Study Background
The study, led by Professor Hayit Greenspan from Tel Aviv University and RADLogics, in collaboration with Dr. Eliot Siegel of the University of Maryland School of Medicine in Baltimore, MD; and Dr. Adam Bernheim of the Icahn School of Medicine at Mount Sinai in New York, NY; found that the CT image analysis algorithm – developed from multiple international datasets – was able to differentiate 157 patients with and without COVID-19 with a 0.996 AUC (plus, 98.2 percent sensitivity and 92.2 percent specificity).
Analyzing Large Numbers of CT Studies for COVID-19
Although it is not recommended as a first-line test, non-contrast thoracic CT has been shown to be an effective tool in the detection, quantification, and follow-up of COVID-19. In addition to detecting and quantifying disease burden, RADLogics’ image analysis further outputs a suggested “Corona Score,” which measures the percentage of lung volume that is infected by the disease.
A consistent and reproducible method for rapidly screening and evaluating high volumes of thoracic CT imaging studies can assist healthcare systems through this pandemic by augmenting radiologists and acute care teams that could be overwhelmed with patients. Additionally, with a greater volume of patients who must be screened for coronavirus, earlier and more rapid detection of positive cases can help improve both the treatment of patients and containment of virus spread.
“This study validates our novel solution, which has been widely studied via multiple international datasets and a range of retrospective experiments to analyze the performance over time,” added Becker. “The conclusion was clear: our rapidly-developed AI-based image analysis can achieve high accuracy in detection of coronavirus as well as quantification and tracking of disease burden.”
Results of this study are available on https://arxiv.org/abs/2003.05037, and it has been submitted to the Radiology Society of North America (RSNA) for review and potential publication in Radiology: Artificial Intelligence. RADLogics is also expanding the initial study to a larger population.
To meet the growing worldwide pandemic, RADLogics also announced that it has rapidly deployed this new CT image analysis algorithm in China, Russia and Italy, and the company is rapidly scaling in other countries in response to the strong demand.
