3D UTE bicomponent imaging of cortical bone using a soft-hard composite pulse for excitation.

PURPOSE To evaluate 3D UTE bicomponent imaging of cortical bone ex vivo and in vivo using a newly designed soft-hard composite pulse for excitation. METHODS Chemical shift artifacts, presenting as fat-water oscillation or combination-induced signal oscillation, significantly reduce the accuracy of quantitative UTE bicomponent analysis of cortical bone. To achieve fat suppression for more reliable bicomponent analysis, a newly developed soft-hard excitation pulse was used with UTE imaging and compared with a single rectangular pulse excitation without and with a conventional fat saturation (FatSat) module. These 3 sequences were applied to 8 bovine bone samples without marrow fat, 3 bovine bone samples with marrow fat, and tibial midshafts of 5 healthy human volunteers. Bicomponent analyses were performed in both ex vivo and in vivo studies. RESULTS The soft-hard pulse provided comparable fat suppression, but much reduced bone signal attenuation compared with the FatSat module. Better bicomponent T2∗ fitting was also achieved with the soft-hard excitation pulse because it greatly reduced chemical shift artifacts and outperformed the single rectangular pulse without or with FatSat. Although the FatSat module reduced fat signals and related fat-water oscillation, the water signals were significantly attenuated with more than 40% reduction due to direction saturation. For the inner layer of tibial midshaft in healthy volunteers, fitting errors increased from 3.78% for the soft-hard pulse to 11.43% and 5.16%, respectively, for the single rectangular pulse without and with the FatSat module. CONCLUSION The 3D UTE sequence with a new soft-hard excitation pulse allows more reliable bicomponent imaging of cortical bone.