Viz LVO stroke triage (Viz.ai)

entry · p2026-09-06-290 · validated

Arrival-to-NIR notification in the AI cohort was: (1) 39.5 min faster (44.13% reduction

tech: learned-lvo-triage
passes the definition as: a deep network fit to labelled CT angiograms that detects a suspected large-vessel occlusion and pages the neurointerventionalist directly, ahead of the radiologist's read
from: 2018-02
region: US
domain: health
side: benefit
people: 14,116 acute stroke alerts (proxy for people) (2022-03)
effect: 39.5 minutes (2022-03) — faster from patient arrival to first contact with the interventionalist about thrombectomy at facilities running the AI than at facilities without it, a 44.13 percent reduction, in a window where treatment delay is the main determinant of disability
reversible: yes
who holds the switch: owner:Viz.ai
attribution: named
source kind: primary
the statistic: VALIDATE (Journal of Stroke and Cerebrovascular Diseases / PMC12802620): 14,116 acute stroke consultations at 166 facilities in 17 states, 1 December 2021 to 31 March 2022, 8,557 at AI facilities and 5,559 at non-AI facilities. Arrival-to-interventionalist-notification was 39.5 minutes faster overall (44.13 percent), 33.0 minutes faster at non-thrombectomy centres and 34.0 minutes faster at thrombectomy-capable centres. A single-centre before-and-after study of 188 patients (86 pre-AI, 102 post-AI; https://pmc.ncbi.nlm.nih.gov/articles/PMC12778824/) found door-in to puncture cut by 86.7 minutes (206.6 to 119.9) and significantly higher rates of adequate reperfusion. The unit is a named proxy: the paper counts stroke alerts and consultations, not distinct persons, and its authors state plainly that they could not capture patient outcomes. In severity, minutes to reperfusion in an LVO stroke is the difference between walking out and not, so I would call this severe per person, and absorbing for the subset whose disability it prevents.
source: https://pmc.ncbi.nlm.nih.gov/articles/PMC12802620/
proposed by: researcher (opus), brief 2026-09-06, severe-benefits
received: 2026-09-06T17:40:15-04:00
harvest report: impacts/harvest/2026-09-06-severe-benefits.md

Rulings, in order

2026-09-06T17:40:16-04:00 · validatorpass — quote found on the fetched page (exact match; HTTP 200 text/html; charset=utf-8)
2026-09-06T17:40:46-04:00 · sessionseverity — Forty minutes sooner to the interventionalist in a large-vessel stroke: a treatment in service, and disability prevented for some — severe. The unit is stroke alerts, a proxy the vocabulary does not take; one consultation is one patient presentation, so the reach is derived on the 8,557 consultations at the facilities running the model (Dec 2021–Mar 2022), the paper's own AI cohort.

The proposal as it arrived (JSON)

{
 "id": "p2026-09-06-290",
 "received": "2026-09-06T17:40:15-04:00",
 "proposer": "researcher (opus), brief 2026-09-06, severe-benefits",
 "kind": "entry",
 "sha": "a68f563941087e0a",
 "data": {
  "kind": "entry",
  "tech": "learned-lvo-triage",
  "tech_name": "Viz LVO stroke triage (Viz.ai)",
  "definition": "a deep network fit to labelled CT angiograms that detects a suspected large-vessel occlusion and pages the neurointerventionalist directly, ahead of the radiologist's read",
  "region": "US",
  "domain": "health",
  "side": "benefit",
  "from": "2018-02",
  "people": {
   "n": 14116,
   "unit": "acute stroke alerts (proxy for people)",
   "date": "2022-03"
  },
  "effect": {
   "n": 39.5,
   "unit": "minutes",
   "date": "2022-03",
   "what": "faster from patient arrival to first contact with the interventionalist about thrombectomy at facilities running the AI than at facilities without it, a 44.13 percent reduction, in a window where treatment delay is the main determinant of disability"
  },
  "reversible": true,
  "switch": "owner:Viz.ai",
  "statistic": "VALIDATE (Journal of Stroke and Cerebrovascular Diseases / PMC12802620): 14,116 acute stroke consultations at 166 facilities in 17 states, 1 December 2021 to 31 March 2022, 8,557 at AI facilities and 5,559 at non-AI facilities. Arrival-to-interventionalist-notification was 39.5 minutes faster overall (44.13 percent), 33.0 minutes faster at non-thrombectomy centres and 34.0 minutes faster at thrombectomy-capable centres. A single-centre before-and-after study of 188 patients (86 pre-AI, 102 post-AI; https://pmc.ncbi.nlm.nih.gov/articles/PMC12778824/) found door-in to puncture cut by 86.7 minutes (206.6 to 119.9) and significantly higher rates of adequate reperfusion. The unit is a named proxy: the paper counts stroke alerts and consultations, not distinct persons, and its authors state plainly that they could not capture patient outcomes. In severity, minutes to reperfusion in an LVO stroke is the difference between walking out and not, so I would call this severe per person, and absorbing for the subset whose disability it prevents.",
  "source": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12802620/",
  "quote": "Arrival-to-NIR notification in the AI cohort was: (1) 39.5 min faster (44.13% reduction",
  "attribution": "named",
  "source_kind": "primary"
 },
 "from_file": "impacts/harvest/2026-09-06-severe-benefits.md"
}