Medical Technology · us
From Finding to Tracking Brain Aneurysms: Bracco and Avicenna.AI Software Receives FDA Clearance
Two CTA imaging tools help identify suspected aneurysms and compare follow-up scans, respectively. A publicly available reader study showed improvements in some detection measures, but superiority over reading without AI assistance has not been demonstrated for the primary endpoint.
A tiny bulge in a blood vessel on brain imaging may be discovered incidentally during an examination for another problem. Once confirmed, subsequent imaging raises another question: has it grown? On November 13, Bracco and Avicenna.AI announced that their jointly developed UPTUNITY | Aneurysm had received two 510(k) clearances from the U.S. Food and Drug Administration (FDA), extending AI assistance to the identification and monitoring of brain aneurysms.
The product uses head computed tomography angiography (CTA) images and has two distinct uses. The detection version, UPTUNITY | Aneurysm Detection, registered under the regulatory name CINA-Aneurysm (K262062), assists trained specialist physicians in identifying suspected saccular brain aneurysms larger than 3 millimeters, providing location and size information. The management version, UPTUNITY | Aneurysm Management, registered as CINA-Aneurysm+ (K262147), measures known aneurysms and compares examination results from different time points.
The two tools address different clinical tasks: the former helps physicians find suspicious lesions in images, while the latter converts size changes during follow-up into comparable data. A report published the same day by Diagnostic Imaging also confirmed this division of functions. However, the greater-than-3-millimeter threshold defines the detection tool’s scope of use; it does not mean smaller lesions can be considered ruled out. The software is also not authorized to independently determine a diagnosis or treatment.
The FDA’s publicly available summary for the detection version provides more specific validation data than the press release: eight U.S. board-certified radiologists interpreted the same set of 230 CTA cases with and without software assistance, with four weeks between the two reading sessions. Interpretations by three independent neuroradiology experts established the reference standard. With assistance, patient-level sensitivity increased from 92.2% to 96.2%, while specificity changed from 87.1% to 88.4%.
These improvements, however, need to be considered alongside the primary result. ROC AUC, the study’s primary measure of overall discrimination, increased from 0.9375 to 0.9497, but the difference was not statistically significant (p = 0.075). Superiority on the primary endpoint has therefore not been demonstrated. This reader study supports some detection benefits, but it does not allow direct conclusions that the software can reduce aneurysm rupture or improve patient outcomes. Nor can the detection version’s results serve as evidence of the management version’s performance in follow-up measurements.
The regulatory status also establishes boundaries for use. The clearances granted are medical device 510(k) marketing clearances, centered on a review of substantial equivalence to legally marketed devices. Physicians must still review the original CTA images and complete the patient assessment; they cannot make or confirm a diagnosis solely from the software’s output. According to the product information provided for this announcement, UPTUNITY | Aneurysm has not yet obtained the European Union’s CE mark.
This also marks concrete progress following the two companies’ announcement in March this year of an AI collaboration for contrast-enhanced CT imaging. Imaging Technology News noted that it is the first product in Bracco’s newly established UPTUNITY clinical decision support portfolio. Avicenna.AI will handle promotion and commercial distribution. The companies are preparing to introduce it in the United States and plan to showcase it at the upcoming annual meeting of the Radiological Society of North America. As the product moves from marketing clearance into routine image interpretation, evidence still needs to accumulate on its performance across different hospital settings and on how these imaging prompts affect subsequent care.