Accuracy of 7T high-resolution vessel wall imaging versus CTA in middle cerebral artery stenosis: a DSA-based validation study

To compare the accuracy, agreement, and reproducibility of 7-Tesla high-resolution vessel wall imaging (7T HR-VWI) and computed tomography angiography (CTA) for quantitative assessment of middle cerebral artery (MCA) stenosis using Digital subtraction angiography (DSA) as the reference standard. This retrospective study recruited 115 patients with non-cardioembolic ischemic stroke between September 2022 and September 2025, comprising 122 MCA stenotic lesions. Two blinded radiologists independently measured stenosis severity (based on diameter and area) and stenotic length using CTA and HR-VWI images. The concordance correlation coefficient (CCC) and Bland–Altman analysis (including assessments of systematic bias, proportional bias, and 95% limits of agreement [LoA]) were used to evaluate agreement between CTA/7T HR-VWI and DSA for diameter-based stenosis and lesion length measurements. Spearman correlation analysis was performed to assess the association between CTA/HR-VWI area-based stenosis measurements and DSA diameter-based stenosis measurements. Intraclass correlation coefficients (ICC) were calculated to assess inter- and intra-observer repeatability. The CCCs for diameter-based stenosis measurements were 0.81 for CTA and 0.92 for HR-VWI. Bland–Altman analysis showed mean biases of 0.06 (95% LoA, −0.04 to 0.15) for CTA and 0.02 (95% LoA, −0.07 to 0.10) for HR-VWI. Area-based stenosis measurements were significantly correlated with DSA diameter-based stenosis (ρ = 0.79 for CTA and 0.90 for HR-VWI; both P < 0.01). For lesion length measurements, the CCCs were 0.91 for CTA and 0.94 for HR-VWI, with mean biases of 0.48 mm (95% LoA, −2.35 to 1.40 mm) and 0.03 mm (95% LoA, −1.49 to 1.57 mm), respectively. Inter- and intra-observer ICCs ranged from 0.83 to 0.97 for CTA and 0.94 to 0.99 for HR-VWI. 7T HR-VWI showed excellent agreement with DSA and high measurement reproducibility for assessing MCA stenosis severity and lesion length. In addition, area-based stenosis measurements derived from HR-VWI were strongly associated with DSA diameter-based stenosis, supporting the potential value of area-based assessment in intracranial stenotic lesions.