ABSTRACT
PURPOSE: To investigate the correlation of non-heme iron content in deep gray matter nuclei as a function of age using quantitative susceptibility mapping (QSM) from both whole-structural and regional perspectives. MATERIALS AND METHODS: We studied a group of 174 normal subjects ranging from 20 to 69 years old and measured the magnetic susceptibility of seven subcortical gray matter nuclei. SWI (susceptibility-weighted imaging) phase images were used to generate the susceptibility maps, which were acquired on a 1.5T scanner. The 3D whole-structural measurements were used to determine age-related thresholds, which were applied to calculate the local iron deposition (RII: portion of the structure that contains iron concentration larger than the structure threshold). Age-susceptibility correlation was reported for each measured structure for both the whole-region and two-region (low iron and high iron content regions) analysis. RESULTS: For the local high iron content region, a strong age-susceptibility correlation was found in the caudate nucleus (CN,R = 0.9), putamen (PUT,R = 0.9), red nucleus (RN,R = 0.8), globus pallidus (GP,R = 0.7), substantia nigra (SN,R = 0.5), and pulvinar thalamus (PT,R = 0.5); for the global iron content, a strong age-susceptibility correlation was found in CN(R = 0.6), PUT(R = 0.7), and RN(R = 0.6). Overall, for each structure analyzed in this study, regional analysis showed higher correlation coefficient and higher slope comparing to the whole-region analysis. Further, we found the quantitative conversion factor between magnetic susceptibility and iron concentration to be 1.03 ± 0.03 ppb per µg iron/g wet tissue. CONCLUSION: We conclude that the age-susceptibility correlation can serve as a quantitative magnetic susceptibility baseline as a function of age for monitoring abnormal global and regional iron deposition. A regional analysis has shown a tighter age related behavior, providing a reliable and sensitive reference for what can be considered normal iron content for studies of neurodegenerative diseases. J. Magn. Reson. Imaging 2016;44:59-71.