Diffusion-weighted imaging as a non-gadolinium alternative for immediate assessing nonperfused area of adenomyosis after magnetic resonance-guided focused ultrasound (MRgFUS) ablation: a potential technique but with slightly overestimate

Background: It is a technical challenge to monitor ablation outcome during magnetic resonance-guided focused ultrasound (MRgFUS) treatment using non-gadolinium technique. The study aimed to investigate the value of diffusion-weighted imaging (DWI) for immediately assessing nonperfused area of adenomyosis after MRgFUS treatment. Methods: We retrospectively included patients with adenomyosis who underwent MRgFUS ablation and underwent both DWI (b=800 sec/mm2) and contrast-enhanced (CE) magnetic resonance (MR) imaging within 15 minutes after treatment. Two blinded observers independently reviewed the DWI scan of the ablated necrotic lesions and measured their area in the central slice of DWI and CE imaging. Consistency and differences in the assessment result were compared. Results: A total of 48 women with adenomyosis (mean age 39.6±4.9 years) were analyzed. Abnormal signals were observed in all 48 adenomyosis lesions on DWI images and could be categorized into 3 types: large area of central low-signal with complete (type 1) or incomplete (type 2) high-signal ring, or inhomogeneous high-signal areas without a ring sign (type 3). Intra- and interobserver intraclass correlation coefficients (ICCs) were 0.84 and 0.80, respectively, for categorizing the DWI signal types (both P<0.001). The DWI abnormal signals and nonperfused areas were essentially corresponding. ICCs were ranged from 0.85 to 0.91 for area measurements using DWI vs. CE MR imaging (all P<0.001). The area of necrotic lesions measured using DWI images was larger than that using CE images (17.17±7.79 vs. 15.41±7.36 cm2, P<0.001). Conclusions: DWI can serve as a non-gadolinium technique for the initial evaluation of nonperfused area of adenomyosis after MRgFUS ablation. However, it is also important to note that DWI may slightly overestimate the nonperfused area.