ArticleNeuroradiology2026
Comparison of AI-assisted quantitative collateral score and visual collateral score in stroke thrombectomy triage.
Article in Neuroradiology, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Not yet cited in PubMed.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
0 citing papers in PubMed.
No citing paper in PubMed yet.
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
6 authors.
Funding
Abstract
purposeTo evaluate an artificial intelligence (AI)-assisted quantitative Collateral Score (qCS) from computed tomography angiography (CTA) for selecting mechanical thrombectomy (MT) candidates, comparing its performance to visual collateral score (vCS).
methodsWe retrospectively analyzed 118 acute ischemic stroke patients with large vessel occlusion. Patients were divided into derivation (n = 94) and validation (n = 24) cohorts. All underwent CTA and CT perfusion (CTP). qCS was derived using a U2-Net model, calculated as the vessel volume ratio between affected and healthy hemispheres. vCS was assessed using Tan and Menon scores. Infarct core volume (ICV), hypoperfusion volume (HV), mismatch volume, and mismatch ratio were calculated from CTP images. Correlations between collateral scores and perfusion parameters were analyzed. Receiver Operating Characteristic (ROC) curves determined optimal thresholds for MT selection, and DeLong test was used for performances comparison.
resultsBoth qCS and vCS correlated negatively with ICV and HV (|ρ|=0.38–0.66, P < 0.001) and positively with mismatch ratio (ρ = 0.52–0.58, P < 0.001). Both scores significantly predicted ICV and HV (|β|=0.45–0.70, P < 0.001), but not mismatch volume. qCS showed superior discrimination (AUC = 0.96) compared to the Tan score (AUC = 0.88, P < 0.001). The optimal qCS threshold was 33.70%, yielding sensitivity of 0.96 and specificity of 0.83. In the validation cohort, qCS achieved the highest accuracy (0.88) and sensitivity (0.95), outperforming both Tan (0.75 and 0.75, respectively) and Menon score (0.83 and 0.90, respectively).
conclusionThe AI-assisted qCS is an effective tool for selecting MT candidates and demonstrated better performance than traditional vCS methods, supporting its potential use in acute stroke thrombectomy triage.
Indexed as
Identifiers
41718730What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.