Quantitative susceptibility mapping: Report from the 2016 reconstruction challenge

Abstract

Purpose: The aim of the 2016 quantitative susceptibility mapping (QSM) reconstruction challenge was to test the ability of various QSM algorithms to recover the underlying susceptibility from phase data faithfully. Methods: Gradient-echo images of a healthy volunteer acquired at 3T in a single orientation with 1.06 mm isotropic resolution. A reference susceptibility map was provided, which was computed using the susceptibility tensor imaging algorithm on data acquired at 12 head orientations. Susceptibility maps calculated from the single orientation data were compared against the reference susceptibility map. Deviations were quantified using the following metrics: root mean squared error (RMSE), structure similarity index (SSIM), highfrequency error norm (HFEN), and the error in selected white and gray matter regions. Results: Twenty-seven submissions were evaluated. Most of the best scoring approaches estimated the spatial frequency content in the ill-conditioned domain of the dipole kernel using compressed sensing strategies. The top 10 maps in each category had similar error metrics but substantially different visual appearance. Conclusion: Because QSM algorithms were optimized to minimize error metrics, the resulting susceptibility maps suffered from over-smoothing and conspicuity loss in fine features such as vessels. As such, the challenge highlighted the need for better numerical image quality criteria.

Association of bone mineral density and fat fraction with magnetic susceptibility in inflamed trabecular bone

“…Susceptibility maps were obtained from all multiecho images using the following, optimized QSM pipeline: (1) Three‐point Dixon method to estimate a field map without fat‐water chemical shift effects, (2) Laplacian phase unwrapping to remove temporal and spatial phase aliasing, (3) Projection onto dipole fields to remove background fields, and (4) direct k‐space inversion using Tikhonov regularization to calculate the susceptibility maps.…”

“…The tilt of the coronal slices was accounted for by defining the dipole kernel to be parallel to the real direction of the main magnetic field in steps 3 and 4. The Tikhonov regularization parameter was set to α = 0.05 in step 4 based on the optimized value in Langkammer et al…”

Weak‐harmonic regularization for quantitative susceptibility mapping

“…A noise‐corrupted field map was forward‐simulated (additively with SNR = 40) using (1) the 12‐head‐orientations COSMOS reconstruction included in the 2016 QSM Reconstruction Challenge data set from spoiled 3D‐GRE scans acquired on a 3T Siemens Tim Trio system using a 32‐channel head‐coil with 1.06‐mm isotropic voxels, 15‐fold wave‐CAIPI acceleration, 240 × 196 × 120 matrix size, TE/TR = 25/35 ms, flip angle = 15°; and (2) a synthetic distribution of external sources (Figure ). These external sources were set to have 2 values: 0.0 ppm for simulated tissue, and 0.94 ppm for simulated air, bone, and so on.…”

“…D, Simulated wrapped‐phase distribution. Phase unwrapping results with Laplacian (E) and Laplacian iterative methods (F), as described by Schofield and Zhu…”

“…Solutions from single axial head‐orientation GRE data (same imaging parameters as those described previously) were also evaluated in the context of the 2016 QSM Reconstruction Challenge with respect to χ 33 and COSMOS ground‐truth references using RMSE, HFEN, and SSIM scores. Two algorithm initializations were explored: (1) the Reconstruction Challenge’s local field map (i.e., 2‐step background prefiltering [LBV followed by fourth‐order polynomial fitting]) and (2) field preconditioning with the proposed nPDF method.…”