Abstract
PurposeTo model and correct the dephasing effects in the gradient‐echo signal for arbitrary RF excitation pulses with large flip angles in the presence of macroscopic field variations.MethodsThe dephasing of the spoiled 2D gradient‐echo signal was modeled using a numerical solution of the Bloch equations to calculate the magnitude and phase of the transverse magnetization across the slice profile. Additionally, regional variations of the transmit RF field and slice profile scaling due to macroscopic field gradients were included. Simulations, phantom, and in vivo measurements at 3 T were conducted for R2∗ and myelin water fraction (MWF) mapping.ResultsThe influence of macroscopic field gradients on R2∗ and myelin water fraction estimation can be substantially reduced by applying the proposed model. Moreover, it was shown that the dephasing over time for flip angles of 60° or greater also depends on the polarity of the slice‐selection gradient because of phase variation along the slice profile.ConclusionSubstantial improvements in R2∗ accuracy and myelin water fraction mapping coverage can be achieved using the proposed model if higher flip angles are required. In this context, we demonstrated that the phase along the slice profile and the polarity of the slice‐selection gradient are essential for proper modeling of the gradient‐echo signal in the presence of macroscopic field variations.
Highlights
The sensitivity of gradient-echo (GRE) imaging to variations in magnetic susceptibility has led to a widespread range of applications in MRI
Approaches focusing on the signal decay are methods such as FMRI, in which the susceptibility difference between deoxyhemoglobin and oxyhemoglobin is measured,[1,2] perfusion MRI with gadolinium-based contrast agent,[3] R∗2 mapping by acquiring multi-gradient-echo, or the determination of the myelin water fraction
A challenge with quantifying mGRE data are the macroscopic field variations that arise, for example, from air/ tissue boundaries, which lead to a faster signal decay and mask tissue-relevant mesoscopic-scale and microstructure-scale R∗2 effects.[12]
Summary
The sensitivity of gradient-echo (GRE) imaging to variations in magnetic susceptibility has led to a widespread range of applications in MRI. For R∗2 mapping, subjects were scanned twice with a mGRE sequence with alternating Gslice polarity using a sinc-Hanning-windowed excitation pulse (Tpulse = 2 ms and BWT = 2.7) with α = 85° (Ernst angle assuming T1 = 1 second). Other sequence parameters were as follows: sinc-Hanning-windowed excitation pulse with Tpulse = 1 ms and BWT = 2, α = 85°, Gslice = 14.15 mT/m, FOV = 255 × 105 mm[2], in-plane resolution = 1.14 × 1.14 mm2, ∆z = 4 mm, 27 bipolar echoes with bandwidth = 500 Hz/Px, TE1 = 2.37 ms, ΔTE = 2.2 ms, TR = 2 seconds, TEnavi = 63.8 ms, 25 interleaved slices with 0% interslice gap, and total scan time = 12 minutes. The differences between the pulses were assessed by estimating R∗2 maps with S4
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
