Advantages of a T1-Weighted Gradient-Recalled Echo (GRE) Sequence With a Radial 3D Sampling Approach Versus 2D Turbo Spin-Echo and Cartesian 3D GRE Sequences in Head and Neck MRI

The purpose of this study is to compare the image quality, including artifacts, of a T1-weighted gradient-recalled echo (GRE) MRI sequence with a radial sampling approach to that of both 2D turbo spin-echo (TSE) and cartesian 3D GRE MRI sequences performed in the head and neck region. The retrospective study included 26 datasets of patients who underwent MRI examination for inflammatory or neoplastic diseases of the head and neck region performed using a 1.5-T system. All examination protocols comprised three fat-saturated T1-weighted sequences performed in the axial plane after contrast agent administration. Axial FOV and spatial resolution in plane and along the -axis were recorded. Sequences were evaluated independently by two readers for qualitative and quantitative parameters, including homogeneity of fat saturation and discrimination of pharyngeal wall structures. Qualitative parameters were evaluated using a 5-point Likert scale. For comparison, mean values of Likert scale scores were generated from the two readers' data. Fat saturation was significantly better on the radial GRE sequences (1.942) than on the TSE (2.346; = 0.002) and cartesian GRE (2.365; = 0.008) sequences. The overall impact of foreign material artifacts was comparable among sequences (radial GRE, 1.731; TSE, 1.731 [ = 1.000]; cartesian GRE, 1.769 [ = 0.414]); however, on direct comparison, the mean area that was not evaluable because of susceptibility was smaller on radial GRE sequences (53.8 mm) than on TSE sequences (59.8 mm; = 0.062) and cartesian GRE sequences (59.6 mm; = 0.079). Overall image quality was 1.635 with the use of radial GRE, 2.423 with TSE ( = 0.000), and 2.500 with cartesian GRE ( = 0.000). For imaging of the head and neck region, radial GRE has several advantages, including improved fat saturation and reduced extent of susceptibility artifacts, compared with TSE and cartesian GRE, leading to improved overall image quality.