3D-printed Shepp-Logan phantom as a real-world benchmark for MRI.

PURPOSE As prescribed and reliable geometrical entities, phantoms have served as indispensable validation tools in a variety of MR-related topics. Though a number of phantoms have been conceived, certain applications may warrant highly customized geometries. The purpose of this study was to demonstrate the expediency of rapid prototyping for generating a flexible class of MR-compatible phantom designs. METHODS An incarnation of the three-dimensional Shepp-Logan numerical phantom, amended for use in magnetic resonance spectroscopic imaging, was actualized using rapid prototyping. Each of the comprising compartments was filled with a solution containing prepared concentrations of common (1)H brain metabolites. Analytical Fourier expressions for the phantom class were established in order to generate a set of simulated measurements, which were then contrasted with acquired data. RESULTS Experimental results for both structural and spectroscopic imaging substantiate the suitability of rapid prototyping for MR phantom applications. The analytically simulated measurements show excellent agreement with the measured data, but also highlight the various consequences effectuated when certain aspects of the acquisition model are disregarded or misrepresented. CONCLUSION Rapid prototyping offers a novel and versatile framework for MR phantom-based validation studies. Furthermore, the growing accessibility and open-source compatibility may provide an important link between the often disparate numerical and haptic testing.