Neurology · Non-contrast head CT triage and notification
qER
Qure.ai
qER is one of the most widely validated head-CT triage tools in the field, with a genuinely global regulatory footprint — FDA clearance under product code QAS, EU MDR Class IIb certification, and a Health Canada Class III licence — and independent external validation that vendor-built models rarely attract. The strongest of that evidence is a Swedish Stroke Register cohort of 1,649 haemorrhage patients (Neuroradiology 2023) reporting 97 percent sensitivity and excellent volumetric agreement against manual segmentation, corroborated by independent UK and Australian cohorts. But the foundational Lancet 2018 paper is manufacturer-authored, no randomised controlled trial has yet been published (the UK ACCEPT-AI and AI-REACT studies remain protocols), and the 2025 large-vessel-occlusion clearance rests on the 510(k) summary alone. A wide approval base and independent prospective-grade evidence place it at two marks rather than three.
Performance Metrics
Clinical Evidence
qER's evidence base is unusual for a vendor product in that the strongest studies are independent. The pivotal development-and-validation paper (Chilamkurthy et al., Lancet 2018;392:2388-2396, PMID 30318264) is manufacturer-authored: algorithms were trained on a retrospective dataset of 313,318 head CT scans from roughly 20 centres in India (2011-2017) and externally validated on the separately collected CQ500 dataset, reporting an AUC around 0.94 for intracranial haemorrhage. As a vendor-built retrospective study it establishes plausibility rather than independent real-world performance. The independent corroboration is what carries the catalogue judgment. A Swedish Stroke Register cohort drawn from Region Skane (Lund; Neuroradiology 2023, PMID 36323862) evaluated qER-NCCT on 1,649 intracerebral-haemorrhage patients and reported 97 percent sensitivity for identifying ICH, with excellent volumetric agreement against manual segmentation (ICC 0.96) and good agreement against the ABC/2 estimate (ICC 0.86); the authors note the tool tends to underestimate very large bleeds. A UK retrospective audit across 44 hospitals (BJR Open 2024, PMID 39479271; qER v2.0 EU, n=1,315 NCCT, 8.5 percent ICH prevalence) and an Australian high-volume trauma-centre study with a prospective validation arm (Alfred Health, Melbourne; 2022, PMID 36400834, 2,916 scans audited) both examined the tool in live teleradiology and emergency workflows. A separate strand validates the qER-Quant quantification lineage rather than the triage flag: automated lateral-ventricular and cranial-vault volumetry in 13,851 patients (Maragkos et al., Beth Israel Deaconess / Harvard, World Neurosurgery 2021, PMID 33421645) and acute-subdural-haematoma volume and midline-shift measurement in 1,789 patients (World Neurosurgery 2024, PMID 38519018). These speak to quantification accuracy, not to triage notification, and should not be read across to the CADt claim. Two gaps bound the evidence. No randomised controlled trial of qER has been published as of the review date — the UK ACCEPT-AI and AI-REACT studies are registered protocols whose results are not yet available. And qER-CTA, the 2025 large-vessel-occlusion clearance, has no peer-reviewed validation beyond its 510(k) summary; its performance claims rest entirely on the FDA submission.
| Study | Design | n | Sensitivity | Specificity | AUC | Published |
|---|---|---|---|---|---|---|
| Chilamkurthy S, Ghosh R, Tanamala S, et al. (manufacturer-authored, dev + external validation) | RetrospectiveRetrospective | 491 | — | — | ~0.94 ICH (CQ500 external validation set) | Lancet, 2018; PMID 30318264; trained on 313,318 head CTs |
| Region Skane / Lund — Swedish Stroke Register cohort (independent) | RetrospectiveRetrospective | 1,649 | 97% | — | — | Neuroradiology, 2023; PMID 36323862; ICC 0.96 vs manual volume |
| Alfred Health, Melbourne — high-volume trauma centre (retrospective audit + prospective validation) | ProspectiveProspective | 2,916 | — | — | — | 2022; PMID 36400834; prospective validation arm over one month |
| UK teleradiology retrospective audit, 44 hospitals (independent) | RetrospectiveRetrospective | 1,315 | — | — | — | BJR Open, 2024; PMID 39479271; qER v2.0 EU, 8.5% ICH prevalence |
Clinical Pulse
Verified clinician sentiment
Verified clinician reviews launching · Apply to contribute →
Reviews are signed by clinicians verified against their regulator's public register.
Inside the Auris+ Listing
Six more sections complete this device’s Auris+ Listing.
Inside the algorithm
ProPro unlocks the five-stage editorial walkthrough of how this device's AI reaches a verdict — model architecture, training data, inference latency, and where the clinician stays in the loop.
Decision Ledger
ProPro unlocks a private, cross-vendor log of every case you read on this device — what the AI called, what you concluded, and a one-line reason if you overrode it. Ready for the EU AI Act's deployer logging obligations when they land in 2028.
Clinical Evidence Deep Dive
ProPro unlocks the structured clinical-evidence summary — study count, target patient population, and a tabular accuracy-metrics view drawn from peer-reviewed sources.
Peer-Reviewed Publications
ProPro unlocks the curated peer-reviewed publication list with PubMed cross-links — the citation backbone of every editorial verdict.
Post-Market & Regulatory Conditions
ProPro unlocks the post-market surveillance summary, recall record, and the conditions of approval that bound real-world use.
AI Algorithm Version History
ProPro unlocks the chronological record of algorithm version changes — what changed when, drawn from manufacturer changelogs and regulatory filings.
Regulatory Approvals
K251610
Class II
Class IIb
Safety Record
No qER recalls or FDA safety communications were identified in publicly available sources as of June 2026; the MAUDE adverse-event database could not be queried directly from this environment, so this should be read as "none found in public reporting" rather than an exhaustive audit. The triage product is a notification overlay — it does not alter images, remove studies from the reading queue, or drive treatment — which limits the harm surface. The principal documented risk is over-reliance on a negative output: even at the 97 percent sensitivity reported in the Swedish cohort there is a real false-negative floor, and that study specifically observed underestimation of very large haemorrhages, so the absence of a flag does not substitute for the radiologist's independent read.
Intended Use & Indications
qER is a radiological computer-aided triage and notification (CADt) device cleared in the United States under product code QAS (21 CFR 892.2080). Its primary clearance (K200921) analyses adult non-contrast head CT studies in the background and raises a case-level worklist priority flag when it suspects one of four findings — intracranial haemorrhage, mass effect, midline shift, or cranial fracture. It is a passive, parallel-workflow tool: it runs alongside the standard of care, does not remove a study from the reading queue, does not annotate or alter the image, and does not provide a diagnosis. Procurement reviewers should note two things about the label. First, Qure.ai's consumer-facing qER product page at times frames the device around bleed detection, but the cleared K200921 indications for use cover all four findings above; the four-finding scope is what the FDA cleared and what the manufacturer's own clearance announcement describes as an industry-first four-in-one clearance. Second, the qER name spans a family of clearances: qER-Quant (K211222) is a quantification tool, not a triage flag, and qER-CTA (K251610) is a separate CADt clearance for large-vessel occlusion on CT angiography. In every jurisdiction the device is decision support: the final diagnostic and treatment decision remains with the interpreting clinician, and a published sensitivity ceiling means the absence of a flag is not exclusionary.