RapidAI has announced pivotal results from what it describes as “the most extensive real-world study to date” on artificial intelligence (AI) for aneurysm detection. Presented at the 2025 American Association of Neurological Surgeons (AANS) annual scientific meeting (25–28 April, Boston, USA), the retrospective, single-centre study analysed 11,694 consecutive computed tomography angiography (CTA) scans, demonstrating how RapidAI’s clinically validated platform—Rapid Aneurysm—can enhance diagnostic confidence and consistency, particularly in detecting nuanced or hard-to-visualise intracranial anomalies like aneurysms when used in conjunction with standard clinical interpretation.
“Detecting aneurysms is inherently challenging, especially in a busy practice when they’re subtly positioned,” said Reza Dashti (Stony Brook Medicine, Stony Brook, USA), lead study investigator. “This study shows how AI can augment clinical expertise and meaningfully advance diagnostic accuracy in real-world settings.”
Presented at AANS 2025, with additional analyses available in the abstract, the study is—according to RapidAI—the most extensive retrospective, single-centre AI validation for brain aneurysm detection to date.
In the study, RapidAI helped identify nearly 23% more aneurysms than were noted on the original radiology reports. Those additional detections had a median size of 3.9mm, which the company notes is large enough to warrant intervention in many cases. For those aneurysms ≥3mm in size, RapidAI demonstrated “standout performance”, reinforcing the value of its deep clinical AI in enhancing diagnostic accuracy and consistency while also enhancing clinical workflows. In this subset, the platform achieved a sensitivity of 92.5% and a specificity of 96.4%.
“Given the increasing demands on today’s radiology teams, this study speaks to the potential of AI to offer meaningful support and capability enhancement to the real-world challenges of manual interpretation,” said David Stoffel, chief business officer at RapidAI. “This level of diagnostic aid for radiologists not only improves patient outcomes but also holds the potential to reduce downstream costs associated with missed or delayed aneurysm diagnoses.”
RapidAI states that its aneurysm platform combines 3D visualisation, growth tracking, and automated measurement, providing clinicians with a comprehensive view and confidence in their decisions. With this study, the company claims to have not only demonstrated the value of AI in real-world care but also helped “raise the bar” for how hospitals and health systems can support their teams with tools for smarter, faster diagnostics.
