When someone suffers a large vessel occlusion stroke, the clock matters more than almost anything else in medicine, since every minute of delay to treatment costs brain tissue. New clinical data presented by Viz.ai at the American Heart Association’s International Stroke Conference in 2026 shows its AI platform cut door-in-door-out time, the time needed to evaluate, coordinate, and transfer a stroke patient to a comprehensive stroke center, by 44% at a regional hospital, according to Viz.ai and Endovascular Today.
The study behind the numbers
The research, titled “Bridging the Gap: Leveraging AI Technology & Partnerships to Meet DIDO Benchmarks for LVO Patients,” was led by Caezar G. Jara at Adventist Health + Rideout and evaluated the platform’s real-world impact at a regional primary stroke center operating within a hub-and-spoke referral network, the common arrangement where smaller community hospitals stabilize stroke patients before transferring them to larger comprehensive centers equipped to perform endovascular clot-removal procedures.
Where the time savings came from
According to the study, the largest gains came from faster identification of large vessel occlusions and faster activation of care teams: the time from CT angiography completion to AI detection of the occlusion dropped by 84%, and the time to notify the care team fell from 45 minutes down to just 7 minutes. Those combined reductions in identification and notification time drove the overall 44% cut in door-in-door-out time for LVO patients being transferred out of the regional hospital.
The technology doing the coordinating
The gains were attributed in part to Viz Assist, a tool built to support regional and community hospitals by integrating real-time patient context, automating care-team activation, standardizing transfer protocols, and streamlining communication between hospitals. Rather than simply flagging a stroke on a scan, the system is designed to also manage the logistical handoff, aligning imaging findings with the operational coordination needed to get a patient into an ambulance or helicopter faster.
Why this matters beyond one hospital’s numbers
Large vessel occlusion strokes require a highly time-sensitive treatment called mechanical thrombectomy, typically only available at larger comprehensive stroke centers, meaning patients first evaluated at smaller community hospitals depend entirely on how quickly that transfer process runs. Faster door-in-door-out times have been directly linked in prior stroke research to better long-term functional outcomes, since brain tissue death accelerates the longer a clot blocks blood flow, making even modest percentage improvements in transfer speed clinically meaningful across a hospital network’s full volume of stroke patients.
Balancing the promise with the limits
The reported gains come from a single regional hospital’s real-world experience rather than a multi-site randomized trial, so it remains to be seen how consistently a 44% improvement would replicate across hospitals with different staffing levels, existing transfer protocols, and baseline technology adoption. Viz.ai already holds more than 50 FDA clearances deployed across roughly 1,700 hospitals, giving it a large existing installed base to eventually generate broader multi-site data, but for now this remains a single-site case study rather than a definitive network-wide benchmark.
Part of a broader push to prove real-world AI impact
The Adventist Health + Rideout results were presented alongside other Viz.ai data at the same 2026 conference examining patient outcomes and hospital economics tied to its stroke platform, reflecting a broader shift in how AI imaging vendors are expected to prove their value. Regulatory clearance alone, based on retrospective algorithm performance against a comparator, increasingly is not viewed as sufficient by hospital purchasing committees; vendors are under growing pressure to publish prospective, site-specific operational data showing concrete workflow improvements, not just diagnostic accuracy metrics, before health systems commit to network-wide rollouts. That shift mirrors similar findings from Aidoc’s foundation-model CT triage clearance and Mayo Clinic’s pancreatic cancer detection research, both of which paired algorithm performance data with real clinical workflow considerations rather than presenting accuracy figures in isolation.
What’s next
Viz.ai and its hospital partners are likely to push for similar data collection across more hospitals in hub-and-spoke stroke networks to see whether the Adventist Health + Rideout results generalize. If they do, faster AI-assisted triage and transfer coordination could become a standard expectation for regional stroke networks nationally, rather than a differentiator only available at hospitals that have specifically adopted Viz.ai’s platform.
