3D Whole-Kidney T1 Mapping Using Look-Locker Inversion Recovery in Conjunction With Balanced SSFP Readout and Dictionary Matching.
To develop and validate dual breath-hold Look-Locker inversion recovery (Dual 3DLL) MRI for 3D whole-kidney T1 quantification using dictionary matching.
Dual 3DLL employs two breath-hold Look-Locker prepared acquisitions using nonselective adiabatic inversion in conjunction with 3D balanced SSFP readouts at three inversion times for each breath-hold. Retrospective registration of T1-weighted images was implemented to reduce misalignment between two breath-holds. For T1 quantification, Bloch equation-based dictionary matching was applied. Dual 3DLL was validated in phantoms using inversion recovery spin-echo (IRSE) based T1 mapping as a reference and in 11 healthy subjects, where 2D MOLLI-based renal T1 mapping was used as a reference.
Validation in the phantom showed an agreement between Dual 3DLL and IRSE (precision: coefficient of variation = 2.5%, relative error compared to IRSE: 4.3% ± 2.0%). The human feasibility study demonstrated the clinical applicability of Dual 3DLL. Averaged T1 derived from whole-kidney coverage Dual 3DLL was in accordance with that of 2D MOLLI (renal cortex: T1,Dual 3DLL = 1413.41 ± 114.73 ms, T1,MOLLI = 1394.39 ± 58.01 ms; medulla: T1,Dual 3DLL = 1827.48 ± 108 ms, T1,MOLLI = 1843.67 ± 91.54 ms). The coefficients of variation between T1 obtained from Dual 3DLL and MOLLI were 6.1% for the renal cortex and 4.4% for the medulla.
This study demonstrates the feasibility of 3D whole kidney T1 mapping using dual breath-hold Look-Locker T1-weighting in conjunction with balanced SSFP readouts and dictionary matching, and provides a technical foundation for improving our understanding of renal (patho-)physiology.
Dual 3DLL employs two breath-hold Look-Locker prepared acquisitions using nonselective adiabatic inversion in conjunction with 3D balanced SSFP readouts at three inversion times for each breath-hold. Retrospective registration of T1-weighted images was implemented to reduce misalignment between two breath-holds. For T1 quantification, Bloch equation-based dictionary matching was applied. Dual 3DLL was validated in phantoms using inversion recovery spin-echo (IRSE) based T1 mapping as a reference and in 11 healthy subjects, where 2D MOLLI-based renal T1 mapping was used as a reference.
Validation in the phantom showed an agreement between Dual 3DLL and IRSE (precision: coefficient of variation = 2.5%, relative error compared to IRSE: 4.3% ± 2.0%). The human feasibility study demonstrated the clinical applicability of Dual 3DLL. Averaged T1 derived from whole-kidney coverage Dual 3DLL was in accordance with that of 2D MOLLI (renal cortex: T1,Dual 3DLL = 1413.41 ± 114.73 ms, T1,MOLLI = 1394.39 ± 58.01 ms; medulla: T1,Dual 3DLL = 1827.48 ± 108 ms, T1,MOLLI = 1843.67 ± 91.54 ms). The coefficients of variation between T1 obtained from Dual 3DLL and MOLLI were 6.1% for the renal cortex and 4.4% for the medulla.
This study demonstrates the feasibility of 3D whole kidney T1 mapping using dual breath-hold Look-Locker T1-weighting in conjunction with balanced SSFP readouts and dictionary matching, and provides a technical foundation for improving our understanding of renal (patho-)physiology.
Authors
Zhang Zhang, Chen Chen, Tao Tao, Li Li, Wang Wang, Zhang Zhang, Xie Xie, Ma Ma, Niendorf Niendorf, Feng Feng
View on Pubmed