A variable flip angle golden‐angle‐ordered 3D stack‐of‐radial MRI technique for simultaneous proton resonant frequency shift and T1‐based thermometry

• Published in: Magnetic resonance in medicine Vol. 82; no. 6; pp. 2062 - 2076
• Main Authors: Zhang, Le, Armstrong, Tess, Li, Xinzhou, Wu, Holden H.
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc 01.12.2019
• Subjects: Ablation; Adipose tissue; Brain; Breast; Coefficient of variation; Frequency shift; Image contrast; Image filters; Magnetic resonance imaging; Mapping; Masks; MR thermometry; Muscles; Prostate; Proton resonance; proton resonant frequency shift; radial MRI; Resonant frequencies; Stability analysis; T1 mapping; Temperature effects; Temperature probes; Temporal resolution; Thermometry; Ultrasound
• DOI: 10.1002/mrm.27883
• ISSN: 0740-3194; 1522-2594

Purpose To develop and evaluate a variable‐flip‐angle golden‐angle‐ordered 3D stack‐of‐radial MRI technique for simultaneous proton resonance frequency shift (PRF) and T1‐based thermometry in aqueous and adipose tissues, respectively. Methods The proposed technique acquires multiecho radial k‐space data in segments with alternating flip angles to measure 3D temperature maps dynamically on the basis of PRF and T1. A sliding‐window k‐space weighted image contrast filter is used to increase temporal resolution. PRF is measured in aqueous tissues and T1 in adipose tissues using fat/water masks. The accuracy for T1 quantification was evaluated in a reference T1/T2 phantom. In vivo nonheating experiments were conducted in healthy subjects to evaluate the stability of PRF and T1 in the brain, prostate, and breast. The proposed technique was used to monitor high‐intensity focused ultrasound (HIFU) ablation in ex vivo porcine fat/muscle tissues and compared to temperature probe readings. Results The proposed technique achieved 3D coverage with 1.1‐mm to 1.3‐mm in‐plane resolution and 2‐s to 5‐s temporal resolution. During 20 to 30 min of nonheating in vivo scans, the temporal coefficient of variation for T1 was <5% in the brain, prostate, and breast fatty tissues, while the standard deviation of relative PRF temperature change was within 3°C in aqueous tissues. During ex vivo HIFU ablation, the temperatures measured by PRF and T1 were consistent with temperature probe readings, with an absolute mean difference within 2°C. Conclusion The proposed technique achieves simultaneous PRF and T1‐based dynamic 3D MR temperature mapping in aqueous and adipose tissues. It may be used to improve MRI‐guided thermal procedures.