New real-world data presented at this year's European Stroke Organisation Conference (ESOC) suggests that RapidAI’s software more accurately identifies medium vessel occlusions (MeVOs) than Viz.ai in stroke patients.

The San Mateo, California-based company analyzed 1,591 consecutive stroke cases from a comprehensive stroke center. Of the 1,122 cases eligible for evaluation, RapidAI’s CT perfusion imaging identified 93% of MeVOs, compared to 70% detected by Viz.ai.

“CT Perfusion is a powerful tool, but its full value is only realized when paired with high-performing software, especially in complex cases with smaller occlusions like MeVOs,” said Dr. Harmeet Sachdev, lead investigator and neurologist. “RapidAI consistently identified more occlusions than Viz in this study, echoing what we saw in the DUEL study with large vessel occlusions.”
MeVOs involve smaller arteries such as the M2/M3 branches of the middle cerebral artery or A1/A2 and P1/P2 segments of the anterior and posterior cerebral arteries. These occlusions are more difficult to detect and, if missed, may delay intervention in time-sensitive stroke care.

The study, titled AI Detection of Medium Vessel Occlusions: Evaluating Performance of RapidAI vs Viz.ai CT Perfusion in 1,591 Consecutive Code Strokes, underscores the challenges in identifying MeVOs and the role of advanced imaging in supporting clinical decisions.

RapidAI’s perfusion imaging tools have previously been validated in large-scale trials, including DEFUSE 2, SWIFT PRIME, EXTEND-IA, and DEFUSE 3. It remains the only FDA-cleared software indicated for selecting patients for mechanical thrombectomy based on perfusion imaging.

“These real-world clinical findings validate what sets RapidAI apart,” said Dr. David Stoffel, chief business officer at RapidAI. “Accurate perfusion imaging is essential for identifying subtle but serious strokes.”

The ESOC presentation adds to ongoing discussion around the importance of imaging software performance in acute stroke workflows.

Stroke Homepage