Method for dixon MRI, multi-contrast imaging and multi-parametric mapping with a single multi-echo gradient-recalled echo acquisition
View Patent ↗To perform Dixon MRI, generate multi-contrast images, and extract multi-parametric maps, this invention presents a multi-echo gradient echo protocol with two sets of echo trains. An example implementation of the invention at 3 T acquires a short-TE train (ΔTE˜1.2 ms, TE<10 ms), which is used to map B0 inhomogeneity and proton density fat fraction (FF), and a second—susceptibility sensitive—long-TE train (16 ms<TE<45 ms) will enable quantification of local frequency shift (LFS) and susceptibility. The presented pipeline automatically generates co-registered images and maps with/without fat-suppressed, including magnitude- and complex-based FF map, B0 map, anatomical images, brain mask, R2* map, unwrapped phase maps for each echo, susceptibility-sensitive images (SWI, LFS and quantitative susceptibility) for each echo, mean susceptibility-sensitive images for each echo-train. The invention is directly applicable to whole head/neck, liver, knee or even whole body scans with sliding table.
1. A method for Dixon MRI, multi-contrast imaging and multi-parametric mapping with a single multi-echo gradient echo (GRE) acquisition, comprising:
a) performing unipolar or bipolar multi-echo GRE acquisition including a short echo time (TE) train and a long TE train;
b) performing phase error correction associated with the bipolar acquisition over all echoes if bipolar acquisition is employed;
c) performing Dixon MRI for the short-TE train data on a magnitude-based FF proton density fat fraction (FF) map, an R2* map, a B0 field (Δf B0 ) map and a complex-based FF map;
d) performing averaging over the magnitude images for the short-TE train to generate an anatomical image;
e) processing the magnitude signal of all echoes to generate an R2* map;
f) performing phase unwrapping of each individual echo except the first echo;
g) calculating susceptibility weighted imaging (SWI) on an echo-by echo basis based on the unwrapped phase;
h) performing local frequency shift (LFS) mapping;
i) performing quantitative susceptibility mapping (QSM);
j) performing averaging of the SWI, LFS and QSM over the short-TE train and the long-TE train separately;
k) performing fat suppression to the anatomical images, SWI images, LFS and QSM maps.
2. The step of claim 1 wherein performing multi-echo GRE acquisition comprises:
a) selecting TEs optimized for Dixon MRI with a short first echo time and tight echo spacing (the short-TE train);
b) selecting TEs optimized for susceptibility mapping while keeping fat and water approximately in-phase based on a multi-peak fat model (the long-TE train).
3. The step of claim 1 wherein performing phase-error correction comprises:
a) constructing a complex data set using Hermitian product;
b) performing 3D phase unwrapping of the reconstructed complex data set;
c) performing polynomial fitting of the unwrapped phase;
d) performing phase-error correction to the odd and even echoes separately.
4. The step of claim 1 wherein performing Dixon Mill for the short-TE train data comprises:
a) generating the magnitude-based FF and R2* maps using magnitude images based on a multi-peak fat model;
b) generating a B0 field map and a proton density fat fraction (FF) map using complex data based on a multi-peak fat model.
5. The step of claim 1 wherein generating anatomical images comprises:
a) averaging the magnitude images of the short-TE train.
6. The step of claim 1 wherein performing R2* mapping using all echoes comprises:
a) correcting the magnitude images for fat-effects (using the FF) based on a multi-peak fat model;
b) fitting the corrected magnitude images with a single-exponential curve.
7. The step of claim 1 wherein performing phase unwrapping comprises:
a) calculating the Hermitian product between the later and the first echoes;
b) removing the phase term related to (Δf B0 ), resulting in an intermediate complex data set;
c) performing 3D phase unwrapping to the intermediate complex data set;
d) generating the final unwrapped phase by summing up the unwrapped phase for the intermediate complex data set and the Δf B0 -related phase, followed by performing deblurring.
8. The step of claim 1 wherein performing SWI comprises:
a) performing high-pass filtering of the unwrapped phase;
b) generating a phase mask from the filtered phase;
c) multiplying the magnitude images by the phase mask multiple times;
d) performing the above procedures on an echo-by-echo basis.
9. The step of claim 1 wherein performing LFS mapping comprises:
a) generating a brain mask from the magnitude-based FF maps;
b) performing background phase removal
c) generating the final LFS map by converting phase to frequency;
d) performing the above procedures on an echo-by-echo basis.
10. The step of claim 1 wherein performing QSM comprises:
a) performing dipole inversion of the LFS maps;
b) performing the above procedure on an echo-by-echo basis.
11. The step of claim 1 wherein performing fat suppression comprises:
a) multiplying the generated images and maps by (1−FF).
12. The step of claim 9 wherein generating a brain mask comprises:
a) performing 3D Guassian filtering to the magnitude-based FF maps;
b) thresholding the filtered FF maps.