Optimized High-Resolution Mapping of Magnetization Transfer (MT) at 3 Tesla for Direct Visualization of Substructures of the Human Thalamus in Clinically Feasible Measurement Time

Purpose: To optimize contrast-to-noise and spatial resolution of a FLASH-based magnetization transfer (MT) protocol for visualization of substructures in human thalamus. Materials and Methods: Healthy adults Were examined at. 3 Tesla with a three-dimensional (3D) spoiled gradient-echo sequence. The signal-to-noise ratio (SNR) was increased by averaging eight bipolar echo acquisitions (mean echo time = 12.3 ms; bandwidth = 370 Hz/pixel). Three isotropic datasets with different weighting (proton density: flip angle/repetition time = 7 degrees/30 ms; T-1: 20 degrees/30 ms and MT: 10 degrees/48 ms, Gaussian MT prepulse) yielded maps of T-1. signal amplitude, MT ratio and MT saturation for comparison to MP-RAGE images. Measuring time was 23 min using partial k-space acquisition. First, the SNR of MT saturation maps in thalamus was optimized by means of the excitation flip angle. Then. noise and partial volume effects were traded off by means of the resolution. Finally. the contrast within the the thalamus and to adjacent structures was compared between different maps. Results: The optimized MT saturation maps at 0.95 mm isotropic resolution provided the highest contrast. It was most prominent between structures of high axonal content (internal medullary lamina, ventral nuclei) and those containing predominantly neuronal somata (pulvinar. mediodorsal thalamus, geniculate bodies). Conclusion: Semiquantitative MT saturation maps provide an enhanced intra-thalamic contrast. The borders and nuclear groups of the thalamus are reliably delineated; individual assignment of singular nuclei seems feasible.