Objectives: In ultra-short echo time (UTE) imaging, the transverse magnetization experiences significant decay during data acquisition, which gives rise to a loss of signal intensity and spatial resolution in conventionally reconstructed images. The present work proposes an iterative algorithm to correct these adverse effects.
Materials and methods: The algorithm involves solving a large linear system of equations and requires a separate reference scan to map the spatially variant transverse relaxation time. It was implemented and applied to simulations and to experiments with custom-built resolution phantoms. The evaluation focused on its ability to improve the actual point-spread function (PSF) and on its influence upon the local signal-to-noise ratio (SNR).
Results: The algorithm is demonstrated to virtually restore the ideal PSF of the acquisition. It proves to provide images with better signal intensity and spatial resolution but reduced SNR, if the transverse relaxation time is known with sufficient accuracy.
Conclusions: The present work shows the basic feasibility of correcting signal decay effects in UTE imaging.