Fast and Accurate Abdominal PDFF and R2 * Mapping With Model‐Fitted Flip Angle Modulation and Simultaneous Multi‐Slice ( SMS ) 2D Imaging

Purpose To establish and validate an online model fitting‐based flip angle calculation method for flip angle modulated (FAM) chemical shift‐encoded (CSE) MRI integrated with simultaneous multi‐slice (SMS) imaging for achieving fast and accurate abdominal proton‐density fat fraction (PDFF) and R2* quantification. Methods Gaussian functions were fitted to directly determine optimal FA profiles, thereby avoiding computationally intensive iterative optimization. Inter‐slice SMS acceleration was incorporated to improve acquisition efficiency. Fat‐water phantom and volunteer with varying liver fat deposition were scanned at 3.0 T to validate the accuracy of PDFF and R2* quantification across different PE steps with different acceleration methods by inter‐slice SMS (FAM‐SMS), in‐plane GRAPPA (FAM‐GRAPPA), and combination of SMS and GRAPPA (FAM‐Hybrid), using low‐FA 3D acquisitions as reference. Relative noise level (rNL) was defined to evaluate the acquisition efficiency. Results In the phantom study (PDFF: 6.2%–35.7%; R2*: 29.2–154.3 s −1 ), the model‐fitted FA profiles closely matched optimized profiles, showing unbiased PDFF and R2* ( p > 0.05). FAM‐SMS showed high correlation with 3D reference (PDFF: 0.994–0.995, R2*: 0.978–0.988). In vivo study, FAM‐SMS achieved comparable noise levels to 3D acquisition (PDFF: 0.9%–28.7%; R2*: 32.7–85.4 s −1 ), demonstrating 18.7% lower PDFF SD and 18.5% lower R2* SD compared to FAM‐GRAPPA ( p < 0.05). Despite higher SD, FAM‐Hybrid reduced rNL by 8.0% (PDFF) and 17.0% (R2*) versus FAM‐GRAPPA. Conclusion The proposed 2D FAM method with model‐fitted FA calculation and SMS acceleration provides fast, accurate, and motion‐robust liver PDFF and R2* imaging, with superior noise performance to GRAPPA, enabling simplified protocol design.