High-Resolution 3T MR Angiography of the Carotid Arteries: Comparison of Manual and Semiautomated Quantification of Stenosis

• Published in: American journal of neuroradiology : AJNR Vol. 30; no. 1; pp. 46 - 52
• Main Authors: Habibi, R, Lell, M.M, Steiner, R, Ruehm, S.G, Sayre, J.W, Nael, K, Finn, J.P
• Format: Journal Article
• Language: English
• Published: Oak Brook, IL Am Soc Neuroradiology 01.01.2009; American Society of Neuroradiology
• Subjects: Adult; Aged; Algorithms; Artificial Intelligence; Biological and medical sciences; Carotid Arteries - pathology; Carotid Stenosis - pathology; Electrocardiography. Vectocardiography; Electrodiagnosis. Electric activity recording; Female; Head & Neck; Humans; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Investigative techniques, diagnostic techniques (general aspects); Magnetic Resonance Angiography - methods; Male; Medical sciences; Middle Aged; Nervous system; Pattern Recognition, Automated - methods; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Reproducibility of Results; Sensitivity and Specificity; High resolution; Nervous system diseases; Radiodiagnosis; Blood vessel; Carotid; Manuals; Stenosis; Circulatory system; NMR angiography; Artery; Comparative study
• ISSN: 0195-6108; 1936-959X
• DOI: 10.3174/ajnr.A1302

High-resolution contrast-enhanced MR angiography (CE-MRA) acquired at 3T exquisitely depict carotid artery (CA) stenosis. In this study, we examined the agreement of different vessel-analysis tools with manual quantitative measurement by 2 readers using CE-MRA data. Three vessel tools determining the trajectory of the vessel of interest and, subsequently, the vessel dimensions were tested against manual measurements. Diameter and area stenoses were calculated. CE-MRA data of 32 patients with CA stenosis were evaluated. The agreement between different measurements was assessed with kappa statistics after categorizing stenosis (<25%, 25%-49%, 50%-69%, 70%-99%, and 100%). The mean grades of stenosis based on diameter measurements were 59% (readers) and 60%/56%/59% based on the analysis with tools A/B/C (P = 0.2-0.7). kappa values for agreement between readers and the vessel tools were 0.73/0.77/0.77 (tools A/B/C for all CAs) and 0.66/0.74/0.75 (for the symptomatic side). The mean grades of stenoses based on area measurements for tools A/B/C were 68%/63%/69% versus 58% for readers. Values from readers differed significantly from those for tools A and C (P < 0.01). kappa values for agreement between readers and the vessel tools were 0.66/0.55/0.64 (for all CAs) and 0.53/0.44/0.57 (for the symptomatic side). The automated approach allows accurate assessment of vessel dimensions in MRA images at least for diameter measurements according to the North American Symptomatic Carotid Endarterectomy Trial criteria.