P2-249: Using brain 3D-MRI and 1H MRS imaging for early diagnosis of Alzheimer's disease

Tao Wang,Shifu Xiao,Qiu Huang,Kewei Chen,Ning Su,Yuanyuan Liu
DOI: https://doi.org/10.1016/j.jalz.2012.05.2154
2012-01-01
Abstract:There were significant statistical difference in the right-inferior temporal lobe, the right-superior temporal lobe, the right-middle frontal lobe and the left-inferior temporal lobe between aMCI and NC group (P <0.005). The right-inferior temporal had capacity of sensitivity 0.74, specificity 0.65, over-all-accuracy of 0.68 between aMCI and NC. There were significant statistical difference in the left-middle temporal lobe, the left-middle cingulate gyrus, the right-middle frontal lobe and right-medial frontal lobe between AD and NC group (P <0.005). Left hippocampus had the ability to distinguish between AD and NC of sensitivity 0.83, specificity 0.92, over-all-accuracy of 0.89. In the PCG, differences between the test groups were found in NAA/Cr, Cho/Cr, mI/Cr and NAA/mI ratios. Group differences were also found in mI/Cr and NAA/mI ratios in the left hippocampus, and in mI/Cr and NAA/mI ratios in the right hippocampus. In discriminate and ROC (Receiver Operating Characteristic) analyses, a NAA/Cr ratio of ≤ 1.50 in the PCG indicated optimal potential for discriminating between aMCI patients and normal control subjects. aMCI and AD patients had significantly temporal lobe atrophy that compared with the control. The ability to distinguish between AD and NC was similar to most studies. Proton MRS of the PCG using the NAA/Cr ratio as a metabolic marker indicates considerable potential for distinguishing between aMCI and NC subjects.
What problem does this paper attempt to address?