gms | German Medical Science

Objective: Physiological pulsatile motion affects the reliable acquisition of diffusion weighted images (DWI) typically used for diffusion tensor imaging and might affect estimated diffusion tensor (DT) parameters also referred to during fiber tractography. As a reliable visualization of risk structures such as the corticospinal tract during brain tumor surgery allows maximizing extent of resection, uncertainly in fiber tract reconstruction might arise from pulsatile motion artifacts. Therefore, a compensation of pulsatile motion artifacts using e.g. cardiac gated imaging, leading to prolonged acquisition time for the patients, might be necessary to overcome this uncertainty.

Methods: In order to analyze the effects of pulsatile motion on diffusion tensor parameters 22 healthy volunteers (mean age: 27.55±6.1 years, male/female: 9/13) were included in a study acquiring non-gated and cardiac-gated DWI data at a 3T MRI system (Trio, Siemens, Germany) with 30 diffusion encoding gradients (b=1000s/mm²) and 7 non-weighted images (b=0s/mm²) and a number of excitations of 5. First, we visually inspected the DWI images for obvious pulsatile motion artifacts and calculated standard deviation (SD) maps across DWI data. Second, a Tract Based Spatial Statistics (TBSS) analysis was conducted for DT parameters FA, MD, and the three tensor eigenvalues L1, L2, L3.

Results: Obvious artifacts were found in the superior cerebellum and midbrain in non-gated DWI with higher SD in the non-gated DWI across these regions group. TBSS analysis showed significant FA differences in the left and right subcortex including the CST (left: 1223 voxels, p=0.008 and right: 1283 voxels, p=0.016), the right posterior limb of internal capsule (PLIC) and midbrain (819 voxels, p=0.008) and left PLIC (215 voxels, p=0.016) and the corpus callosum (CC), anterior limb of internal capsule (ALIC) and subcortex (28663 voxels, p=0.002). Other diffusion parameters such as MD, L1, L2 and L3 revealed corresponding changes the subcortex including the CST, the PLIC and midbrain, the ALIC, CC.

Conclusion: Pulsatile motion showed obvious influence on diffusion tensor parameters in a whole brain voxel-based analysis in the subcortex (including CST), the ALIC and PLIC, the CC and the midbrain. Pulsatile motion thereby might also lead to changes in fiber tractography results based on the estimated diffusion tensor and has to be taken into account when analyzing diffusion tensor parameters.