AZmed, one of the leading AI companies in medical imaging, today announced that it has received two new U.S. Food and Drug Administration (FDA) clearances for its AI tool AZchest. The clearances include applications intended to assist radiologists in the interpretation and detection of chest X-rays for lung nodules and triage capabilities for pneumothorax and pleural effusion.
A standalone study on chest X-rays showed that AZchest has extremely high sensitivity in detecting lung nodules, at 88.47%, associated with a high specificity of 82.94%. In a controlled multi-reader multi-case trial, experienced radiologists interpreted a standardized set of chest X-rays both with and without the assistance of AZchest. When aided by AZchest, the sensitivity increased by 10%, reaching 89.35%. The AUC also increased by 5%.
Regarding the triaging of pneumothorax and pleural effusion, the algorithm also confirmed its high detection performance. The sensitivity and AUC for pneumothorax and pleural effusion were 93.79% / 98.57% and 91.34% / 98.30%, respectively.
These results make AZchest one of the most effective AI tools now available in the United States, offering a quantifiable diagnostic advantage in real clinical practice and supporting the detection of cancer at early stages. The performances also sustain AZmed’s scientific reputation and clinical integrity in the medical imaging industry, and this validated capability supports the potential of AZchest to help reduce workload on radiologists.
“We are proud to add two new FDA clearances to our portfolio. Our deep learning algorithms are designed to rapidly and accurately detect abnormalities, thereby ensuring that critical cases are flagged for clinical review promptly,” said Julien VIDAL, CEO of AZmed.
AZmed’s products are used in 2,500+ centers across 55 countries. With AZchest now FDA-cleared for U.S. use, healthcare providers gain access to a mature, proven solution. The achievement builds on the company’s previous success in obtaining FDA clearance for the pediatric indication of its fracture detection AI, AZtrauma, less than a year ago.