Chemical shift imaging in the head and neck at 3T: initial results.

PURPOSE To determine the optimal method to correct air and tissue susceptibility differences in the head and neck to allow proton ((1)H) chemical shift imaging (CSI) to be performed at 3T. MATERIALS AND METHODS Shimming protocols (iterative, first-order, and second-order) and perfluorocarbon (PFC) pads were evaluated using water peak linewidth measurements obtained from single-voxel magnetic resonance spectroscopy (MRS) on a head and neck phantom. After optimization of the technique, CSI was then tested on 14 patients with head and neck tumors. RESULTS Second-order shimming (water peak linewidth, 4.6 Hz) performed significantly (P < 0.001) better than first-order (16.5 Hz) and iterative shimming (18 Hz) and the water peak linewidth was significantly reduced using PFC pads (P < 0.001). Using second-order shimming and PFC pads, CSI was successful in 10 patients with nodal metastases (n = 8) and benign tumors (n = 2) and unsuccessful in four patients with primary tumors along the aerodigestive tract. CONCLUSION Proton CSI can be successfully performed in the head and neck using second-order shimming and PFC pads to correct air and tissue susceptibility differences. CSI was more successfully performed on nodal metastases, while CSI for primary tumors along the aerodigestive tract remains a challenge.