Tool pinpoints incidental findings on CT scans that contain the lung or part of the lung.
Artificial intelligence (AI) radiology solution company Aidoc announced Friday it has received permission from the U.S. Food & Drug Administration (FDA) for its tool designed for the detection of COVID-19 findings.
According to a company statement, the tool detects and prioritizes incidental CT findings, such as ground-glass opacities, in any scans that contain the lung or part of the lungs, including chest, abdomen, and cervical spine. Pinpointing these findings can provide additional information that can, potentially, augment patient evaluation.
The tool is the most recent addition to Aidoc’s AI tools that detect intracranial hemorrhage, large-vessel occlusion, pulmonary embolism in pulmonary angiography, and cervical spine features.
“In our experience, it is not unusual for the radiologist to be the first to diagnose COVID-19 disease in patients especially when the disease is clinically unsuspected. The outbreak of the COVID-19 pandemic may occur in waves and should these waves occur, it will become increasingly important to identify imaging findings suggestive of COVID-19 in a variety of clinical settings,” said Paul Chang, M.D. professor of radiology and vice chair of radiology informatics at the University of Chicago Medicine. “Aidoc’s ability to detect and triage patients with incidental findings associated with COVID-19 acts as another layer of protection as the disease may continue to circulate in the months to come.”
Based on studies from universities in Brussels, Chicago, Brescia, and Maimonides Medical Center in New York, radiologists reviewing CT scans for other reasons, such as oncology or abdominal pain screenings, discovered that 8 percent-to-10 percent of patients were also COVID-19-positive. These patients had not shown any other COVID-19-associated respiratory symptoms.
The goal, company officials said, is for the tool to create a safety net that can help radiologists identify and address potentially life-threatening conditions as they work through the anticipated backlog of imaging studies created by the pandemic.
Study Assesses Lung CT-Based AI Models for Predicting Interstitial Lung Abnormality
September 6th 2024A machine-learning-based model demonstrated an 87 percent area under the curve and a 90 percent specificity rate for predicting interstitial lung abnormality on CT scans, according to new research.