Abstract
BACKGROUND: Safe and accurate imaging of the peripheral arterial system is important for diagnosis and treatment planning of patients with peripheral artery disease (PAD). PURPOSE: To evaluate image quality and diagnostic performance of unenhanced magnetic resonance angiography (MRA) based on velocity-selective (VS) magnetization preparation (termed VS-MRA). STUDY TYPE: Prospective. POPULATION: Thirty-one symptomatic PAD patients underwent VS-MRA. Twenty-four of them underwent clinical digital subtraction angiography (DSA) examination, 18.8 ± 5.2 days after the MR scans. FIELD STRENGTH/SEQUENCE: 1.5T MRI that included VS-MRA (homemade research sequence) and phase-contrast flow imaging (clinical sequence). ASSESSMENT: Image quality (0: nondiagnostic, 3: excellent) and stenosis severity (0: normal, 3: occlusion) of VS-MRA images were assessed independently by three reviewers. Arterial signal-to-noise-ratio (SNR) and artery-to-muscle contrast-to-noise ratio (CNR) were calculated. STATISTICAL TESTS: The sensitivity and specificity of VS-MRA were calculated for the detection of significant stenosis (>50%) with DSA as the reference standard. Interobserver agreement among the three reviewers was evaluated by using Cohen κ-statistics. RESULTS: The image quality score of VS-MRA was 2.7 ± 0.5 for Reader 1, 2.8 ± 0.5 for Reader 2, and 2.8 ± 0.4 for Reader 3; SNR and CNR were 37.8 ± 12.5 and 30.5 ± 11.8, respectively. Segment-based analysis revealed that VS-MRA had sensitivities of 85.3%, 74.5%, and 78.4%, respectively, for the three reviewers, and specificities of 93.5%, 96.8%, and 95.2%. The interobserver agreement for the stenosis grading was good, as demonstrated by Cohen κ values of 0.76 (Reader 1 vs. Reader 2), 0.82 (Reader 1 vs. Reader 3), and 0.79 (Reader 2 vs. Reader 3). DATA CONCLUSION: Unenhanced VS-MRA allows clear depiction of the peripheral arteries and accurate stenosis grading, as evidenced by high image quality scores and strong agreement with DSA. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;49:744-751.