Abstract Background and aims ASPECTS is used to determine eligibility for mechanical thrombectomy (MT), with current guidelines recommending treatment for patients with ASPECTS ≥6. Its main advantage is rapid assessment using non-contrast CT (NCCT). CT perfusion (CTP) imaging is widely utilised to select patients for MT in those presenting outside the determined treatment window. However, this can delay treatment decisions and may not be necessary particularly in cases with acceptable ASPECTS scores. We aimed to determine whether non-contrast CT ASPECTS assessment by a neuroradiologist and AI software (RAPID) correlates with CTP aspects assessment. Furthermore, do these parameters correlate with a small core infarct. Methods Retrospective review of 65 consecutive stroke referrals to a comprehensive stroke centre that had: 1) CT perfusion imaging; 2) evidence of a proximal occlusion within the anterior circulation; 3) Presenting out of the 6-hour window for MT. Baseline characteristics, radiologist ASPECTS, AI aspects and CTP data were collected. McNemars and Wilcoxon tests were utilised for group differences. Simple linear regression with Fischer’s exact test was used to compare non-contrast ASPECTS with core volume. Results No difference between RAPID, neuroradiologist, and CTP ASPECT scores at assigning patients a good (≥7) vs poor (7) score. Higher RAPID and radiologist ASPECTS correlates with lower core volume on CTP (r2=0.59; P0.0001). Radiologist ASPECTS exhibits impressive PPV (92%; CI 0.81-0.97), but low NPV (40%; CI 0.20-0.64) for predicting patients meeting DEFUSE-3 criteria for MT. Conclusions In patient’s presenting outside of the conventional MT window, those with good ASPECTS on NCCT may proceed to thrombectomy without CTP. Conflict of interest Saad Moughal: nothing to disclose. Umar sabir: nothing to disclose. Rans Nadir: nothing to disclose. Rukhtam Saqib: nothing to disclose.
Building similarity graph...
Analyzing shared references across papers
Loading...
Saad Moughal
Umar Sabir
Rans Nadir
European Stroke Journal
Salford Royal Hospital
Building similarity graph...
Analyzing shared references across papers
Loading...
Moughal et al. (Fri,) studied this question.
www.synapsesocial.com/papers/69fd7f86bfa21ec5bbf07fc0 — DOI: https://doi.org/10.1093/esj/aakag023.393