Accelerated motion-robust abdominal 3D T1ρ mapping using diamond radial sampling

To develop and validate a fast, motion-robust, free-breathing abdominal 3D T1ρ mapping method by combining variable-density diamond radial k-space sampling with fast-MAPSS (magnetization-prepared angle-modulated partitioned-k-space spoiled gradient-echo snapshots).

3D MAPSS T1ρ imaging was performed at 3T using five spin-lock time (TSL) pairs in phantom scans and three TSL pairs in nine healthy volunteers. Phantom experiments compared Cartesian sampling (reference) with stack-of-stars and diamond radial sampling. Volunteer scans utilized stack-of-stars and diamond sampling. T1ρ values were calculated using mono-exponential fitting with MAPSS (five or three TSL pairs) and fast-MAPSS (three TSLs). These values were compared across sampling methods and T1ρ techniques in phantom and in vivo for the liver, pancreas, and muscles.

Diamond sampling achieved a 2.2-fold acceleration compared to stack-of-stars. In phantom experiments, mean T1ρ differences from reference Cartesian MAPSS were -1.93% for stack-of-stars, and -0.09% for diamond sampling. Fast-MAPSS with three TSLs resulted in differences of 0.54% (stack-of-stars) and 1.45% (diamond). In volunteers, a strong correlation (R² ≥ 0.88) was observed between diamond and stack-of-stars MAPSS, and between MAPSS and fast-MAPSS for both sampling methods. Bland-Altman analysis revealed minimal biases (<1.5 ms) between the methods. Paired t-tests indicated no significant T1ρ differences between MAPSS and fast-MAPSS or between sampling methods in volunteers.

Fast-MAPSS with diamond radial k-space sampling enables clinically feasible, motion-robust, free-breathing 3D T1ρ mapping of the abdomen with reduced scan time and quantification accuracy comparable to standard methods.