Purpose: Amyloid-beta Plaque (AP) imaging is a key to cure and understand Alzheimer’s Disease (AD). We used MR-phase information to detect AP in which iron is contained. High sensitivity to magnetic susceptibility of phase information plays important role to detect AP and its location on MRI image. However, phase may detect magnetic sources other than iron of AP due to high sensitivity of magnetic susceptibility, e.g., iron in the cortex due to normal aging. Therefore, we need to discriminate other magnetic sources from iron in AP for accurate detection and quantification of AP. In this study, we suggested a double Gaussian model to discriminate iron in AP from other sources operated on phase data. The aim of this study was to examine and show a performance of the model to detect and quantify AP on mouse brain.
Methods: 7T-MRI (BioSpec 70/20 USR, Bruker Biospin, Germany) with 3D-FLASH was used to obtain phase image of mouse brains (APP23 and wild type (C57BL/6JJcl) aged from 9 to 16 m/o). Scan parameters were TE/TR = 12.8/50 ms, FA = 20 deg., spatial resolution = 0.08 iso-boxcel, NEX = 2-24 and scan duration = 10 min – 3 hrs. Six parameters contained in sum of two Gaussian (magnitude, standard deviation and center of mass of each Gaussian) were decided by fitting to phase data distribution obtained from cortex regions. Since one of the Gaussian distribution represented phase distribution of AP, we could define iron phase corresponding to AP. Detected signal had been enhanced on magnitude image, which was so called Phase Difference Enhanced Imaging (PADRE). We had evaluated signal on PADRE image by comparing with staining image of AP.
Conclusion: The double Gaussian model could correctly discriminated AP from the other signal and noise, resulting in delineating AP signal on PADRE image.