Quantitative susceptibility mapping for investigating subtle susceptibility variations in the human brain ☆ Ferdinand Schweser ⁎, Karsten Sommer, Andreas Deistung, Jürgen Rainer Reichenbach Medical Physics Group, Institute of Diagnostic and Interventional Radiology I, Jena University Hospital — Friedrich Schiller University Jena, Philosophenweg 3, 07743 Jena, Germany abstract article info Article history: Accepted 24 May 2012 Available online 1 June 2012 Keywords: Magnetic susceptibility mapping Tissue susceptibility Phase imaging Homogeneity Enabled Incremental Dipole Inversion COSMOS Quantitative susceptibility mapping (QSM) is a novel magnetic resonance-based technique that determines tissue magnetic susceptibility from measurements of the magnetic field perturbation. Due to the ill-posed nature of this problem, regularization strategies are generally required to reduce streaking artifacts on the computed maps. The present study introduces a new algorithm for calculating the susceptibility distribution utilizing a priori information on its regional homogeneity derived from gradient echo phase images and analyzes the impact of erroneous a priori information on susceptibility map fidelity. The algorithm, Homogeneity Enabled Incremental Dipole Inversion (HEIDI), was investigated with a special focus on the reconstruction of subtle susceptibility variations in a numerical model and in volunteer data and was compared with two recently published ap- proaches, Thresholded K-space Division (TKD) and Morphology Enabled Dipole Inversion (MEDI). HEIDI resulted in susceptibility maps without streaking artifacts and excellent depiction of subtle susceptibility variations in most regions. By investigating HEIDI susceptibility maps acquired with the volunteers' heads in different orien- tations, it was demonstrated that the apparent magnetic susceptibility distribution of human brain tissue consid- erably depends on the direction of the main magnetic field. © 2012 Elsevier Inc. All rights reserved. Introduction Bulk magnetic susceptibility is a fundamental physical tissue proper- ty, which is long known to reflect clinically relevant tissue characteris- tics, such as tissue iron content (Bauman and Harris, 1967; Brittenham et al., 1982; Senftle and Thorpe, 1961) or blood oxygenation level (Ogawa et al., 1990). In magnetic resonance imaging (MRI) the magnetic susceptibility is one of the most fundamental contrast mechanisms as it describes the response of the imaged object to the static main magnetic field. Since the early days of clinical MRI strong interest has existed in quantifying magnetic susceptibility in vivo to obtain additional bio- chemical information or to increase the specificity of imaging studies in various pathologies (Cox et al., 1986; Young et al., 1987). During the last decade, gradient (recalled) echo (GRE) phase imaging (Budde et al., 2011; Deistung et al., 2008; Duyn et al., 2007; Rauscher et al., 2005; Walsh and Wilman, 2011), susceptibility weighted imaging (SWI) (Haacke et al., 2004; Reichenbach and Haacke, 2001; Sehgal et al., 2005) and especially T 2 ⁎ -weighted imaging (Li et al., 2006; Pu et al., 2000) have been used to qualitatively assess magnetic susceptibility variations in tissue. Quantitative susceptibility mapping (QSM) uses small magnetic field variations to compute quantitative maps of the corresponding un- derlying magnetic susceptibility distribution. The method has been demonstrated to reveal anatomical contrast in the brain with unprece- dented quality, depicting various deep gray matter nuclei and their intri- cate substructures (Li et al., 2012a; Schweser et al., 2011a; Wharton and Bowtell, 2010). The magnetic susceptibility contrast presented in these initial studies is believed to be useful, e.g., for preoperative targeting of deep brain nuclei in the context of deep brain stimulation (DBS) (O'Gorman et al., 2010; Schäfer et al., 2010, 2012), characterization of brain lesions (Schweser et al., 2010a), brain iron quantification (Bilgic et al., 2011; Schweser et al., 2011a; Zivadinov et al., 2011), and under- standing of pathology-related susceptibility changes in white matter (Pitt et al., 2010). Despite the excellent depiction of tissue anatomy on susceptibility maps, clinical applicability of QSM seems hitherto limited, because the employed high-precision reconstruction schemes rely on excessive data oversampling involving multiple acquisitions of the sub- jects' head at different orientations with respect to the main magnetic field, also referred to as Calculation Of Susceptibility through Multiple Orientation Sampling (COSMOS) (Liu et al., 2009). Recent results, fur- thermore, suggest that COSMOS maps may suffer from orientation NeuroImage 62 (2012) 2083–2100 Abbreviations: COSMOS, Calculation Of Susceptibility through Multiple Orientation Sampling; GRE, gradient (recalled) echo; HEDI, Homogeneity Enabled Dipole Inver- sion; HEIDI, Homogeneity Enabled Incremental Dipole Inversion; MEDI, Morphology Enabled Dipole Inversion; QSM, quantitative susceptibility mapping; RMSE, root- mean-square error; SWI, susceptibility weighted imaging; TKD, Thresholded K-space Division; TV, total variation; WG, weighted gradient. ☆ Author contributions: F.S., A.D., and J.R.R. designed the research; F.S. and K.S. per- formed the research; F.S., K.S. and A.D. contributed new reagents/analytic tools; F.S. and K.S. analyzed data; F.S. wrote the paper; manuscript drafting and editing were per- formed by all authors. ⁎ Corresponding author. Fax: + 49 3641 935081. E-mail addresses: mail@ferdinand-schweser.de (F. Schweser), kar_sommer@gmx.de (K. Sommer), andreas.deistung@med.uni-jena.de (A. Deistung), juergen.reichenbach@med.uni-jena.de (J.R. Reichenbach). 1053-8119/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2012.05.067 Contents lists available at SciVerse ScienceDirect NeuroImage journal homepage: www.elsevier.com/locate/ynimg