Brain MRI with Quantitative Susceptibility Mapping: Relationship to CT Attenuation Values

Abstract

Background: Quantitative susceptibility mapping (QSM) is used to differentiate between calcification and iron deposits. Few studies have examined the relationship between CT attenuation values and magnetic susceptibility in such materials. Purpose:To assess the relationship among metal concentration, CT attenuation values, and magnetic susceptibility in paramagnetic and diamagnetic phantoms, and the relationship between CT attenuation values and susceptibility in brain structures that have paramagnetic or diamagnetic properties. Materials and Methods:In this retrospective study, CT and MRI with QSM were performed in gadolinium and calcium phantoms, patients, and healthy volunteers between June 2016 and September 2017. In the phantom study, we evaluated correlations among metal concentration, CT attenuation values, and susceptibility. In the human study, Pearson and Spearman correlations were performed to assess the relationship between CT attenuation values and susceptibility in regions of interest placed in the globus pallidus (GP), putamen, caudate nucleus, substantia nigra, red nucleus, dentate nucleus, choroid plexus, and hemorrhagic and calcified lesions.Results: Eighty-four patients (mean age, 64.8 years 6 19.6; 49 women) and 20 healthy volunteers (mean age, 72.0 years 6 7.6; 11 men) were evaluated. In the phantoms, strong linear correlations were identified between gadolinium concentration and CT and MRI QSM values (R 2 = 0.95 and 0.99, respectively; P , .001 for both) and between calcium concentration and CT and MRI QSM values (R 2 = 0.89 [P = .005] and R 2 = 0.98 [P , .001], respectively). In human studies, positive correlations between CT attenuation values and susceptibility were observed in the GP (R 2 = 0.52, P , .001) and in hemorrhagic lesions (R 2 = 0.38, P , .001), and negative correlations were found in the choroid plexus (R 2 = 0.53, P , .001) and in calcified lesions (R 2 = 0.38, P = .009). Conclusion:CT attenuation values showed a positive correlation with susceptibility in the globus pallidus and hemorrhagic lesions and negative correlation in the choroid plexus and calcified lesions.

Paradoxical paramagnetic calcifications in the globus pallidus: An ex vivo MR investigation and histological validation study

“…This would explain the paradoxical paramagnetic lesions in the GP: in the medial aspect of the GP, vascular mineral deposition appears hyperdense on CT with paramagnetic bulk magnetic susceptibility. A previous report 24 also reported two distinct relationship patterns between magnetic susceptibility and CT numbers in different regions of the brain. In calcified tissue, such as the choroid plexus or pineal gland, magnetic susceptibility and CT numbers were inversely correlated.…”

“…However, in the GP, they were positively correlated, similar to the relationship of hemorrhagic lesions. 21,24 This phenomenon might be due to the high concentration of iron in both GP and hemorrhagic lesions, regardless of the presence of diamagnetic calcifications. Interestingly, some vessels only showed iron deposition.…”

“…Some studies have evaluated mineral depositions at GP under normal and pathologic conditions using several MRI techniques including QSM. 13,19,[22][23][24] However, no studies have made a histological validation of these interesting findings in GP; even a previous report discussed them as artifacts. 23 The clarification of the paradoxical paramagnetic sources of GP might improve understanding of QSM findings and result in better clinical interpretation.…”

Increased Brain Iron Deposition in the Putamen in Patients with Type 2 Diabetes Mellitus Detected by Quantitative Susceptibility Mapping

“…“Physiologic calcification” in the basal ganglia (including the putamen) is always accompanied by the deposition of other metals (e.g., zinc, iron, copper, manganese, and aluminum) [ 41 ]. We found that T2DM patients presented obviously higher mean susceptibility values in the putamen than HCs, which may be explained by the phenomenon that although calcium in these areas causes negative susceptibility on QSM, this may be overwhelmed by the strong paramagnetic deposits, such as iron, because QSM measures overall bulk magnetic susceptibility [ 42 ]. Meanwhile, microbleeds change local magnetic susceptibility because of the pathologic iron accumulation [ 43 ].…”

Phenotypic and genetic associations of quantitative magnetic susceptibility in UK Biobank brain imaging

“…For example, χ has been demonstrated to have a strong, positive linear relationship with iron concentration (measured using mass spectrometry) in post-mortem brain tissues 26 . χ values in white matter and calcified lesions have also been validated using myelin staining intensity 75 and CT attenuation values 76 . In our reproducibility analysis using the two-timepoint data from 1,447 subjects (Supplementary Fig.…”