Non-parametric intravoxel incoherent motion analysis of the thyroid gland

To implement a protocol for intravoxel incoherent motion (IVIM) of the thyroid, to determine base parameters in healthy volunteers, and to provide preliminary experience on clinical applicability in one patient.

Eight healthy volunteers underwent 3T MRI using a diffusion weighted echo-planar imaging sequence with 12 different b-values between 0-800 s/mm². The IVIM parameters diffusion coefficient D, pseudo-diffusion coefficient D*, perfusion fraction F_p, and the optimal b-values thresholds were calculated for each thyroid lobe, muscle tissue and the cerebrospinal fluid (CSF) using a non-parametric multi-step algorithm and compared with a Student's t-test. A p-value <0.05 was considered significant.

Mean values for healthy thyroid tissue were: D 1.01 ± 0.13 × 10^-3 mm²/s, D* 71.0 ± 52.5 × 10^-3 mm²/s and F_p 17.1 ± 4.2%; for muscle: D 0.50 ± 0.21 × 10^-3 mm²/s, D* 58.3 ± 99.2 × 10^-3 mm²/s and F_p 26.5 ± 9.3%; and for CSF D 2.18 ± 0.93 × 10^-3 mm²/s, D* 99.2 ± 41.2 × 10^-3 mm²/s and F_p 74.6 ± 12.7%. The optimal b-value threshold separating diffusion and perfusion effects in thyroid ranged between 0-70 s/mm². Healthy thyroid tissue showed similar F_p compared to muscle, both lower than CSF.

The proposed IVIM protocol provides surrogate markers on cellular diffusion restriction and perfusion; thereby providing a more comprehensive description of tissue properties compared to conventional DWI.