Quantitative Comparison of Susceptibility-weighted Imaging Methods for Detection of Differences in Deep Grey Matter in Multiple Sclerosis

Abstract

Problem: Investigating signal changes in deep gray matter (DGM) structures isa novel approach to understanding the role of iron in the progression of multiplesclerosis (MS)1,2,3. T2* weighted angiography (SWAN) is a new susceptibility weighted3D imaging method offering less noise than the traditional T2* weighted gradientrecalled echo (T2* GRE) method4. We assess the difference between SWAN and T2*GRE and their ability to detect signal changes in DGM structures and image quality.Method: Five healthy controls and 13 MS patients were selected from an ongoingstudy. MRI was performed on a 3T MR scanner using the standard protocols forthe following sequences: a 3D multi-echo SWAN, a 2D single-echo T2* GRE. Signalmeasurements were taken in DGM structures with the FMRIB software library andcontrast-noise (CNR) and signal-noise (SNR) ratios were calculated. Statistical analysiswas conducted with SPSS v19 using two-way ANOVA and post-hoc for weighted andun-weighted means.Results: An interaction effect was observed between region and module. Controlsconsistently had a higher signal than MS patients in T2* GRE, however this onlyoccurred in two out of the four regions in SWAN. SWAN demonstrated a higher SNRthan T2* GRE offering a cleaner image. T2* GRE and SWAN offered equal contrast ontwo structures. T2* GRE was considerably superior to SWAN in the remaining two.Conclusions: SWAN allows for a cleaner image which may provide a qualitativeadvantage for trained radiologists but is not significantly superior to T2* GRE quantitatively.T2* GRE offered superior contrast to SWAN, providing better separation oftissue. Larger differences in signal intensity between control and MS patients observedin T2* GRE make MS patients more distinguishable. Consistent pattern (control higherthan MS) in T2* GRE makes it more reliable than SWAN for detecting changes in DGMstructures.