Alzheimer's disease (AD) is a debilitating neurodegenerative disease that imposes tremendous burdens to the society today. Research has been undergoing actively to discover new drugs that can alleviate symptoms or slow down the progression of the disease effectively. To screen drugs more effectively, it is desirable to monitor the effect of medication on brain structure and function which precedes behavioral improvement. In this study, a novel neuroprotective drug was administered to APP/PS1 AD mice for two months. The tract integrity was monitored over time using diffusion spectrum imaging (DSI). Three groups of mice were studied, AD mice with treatment or without treatment, and wild type healthy mice. We hypothesized that DSI was able to identify brain regions that differentiated the AD and healthy mice, and showed the drug effects in the AD mice with medication. Subjects: Nine APP/PS1 mice containing human transgenes of AD were used in our experimental mice model. A neuroprotective drug was administered orally to 4 AD mice for two months. For comparison, 5 wild type healthy mice were included as the control group. Imaging: All mice received T1-weighted imaging and DSI on a 7T MRI system (BioSpec, Bruker, USA). The parameters were as follow: T1-weighted: RARE T1 sequence, TR/TE=1300/8 ms, matrix size=256x256, FOV=20x20 mm^2, flip angle=180°, slice thickness=0.4 mm, slice number=15 in trans-axial plane, and 4 averages; DSI: echo planer imaging sequence, 102 diffusion encoding gradients with bmax=4500 s/mm^2, TR/TE=3750/40 ms, matrix size=50x50, FOV=20x20 mm^2, slice thickness=0.4 mm, and slice number=15 in trans-axial plane. Analysis: We used ROI-base analysis method to estimate the generalized fractional anisotropy (GFA, a measure of tract integrity) values in these three groups. The procedures of ROI-base analysis method were described as follow. 1) Null images of the mice brains were registered to create a study specific template (SST) using SPM12. 2) Several regions, including corpus callosum (CC), hippocampus (Hp), fimbria (fi), thalamus (Th) and internal capsule (IC), were selected as ROIs in the SST space. 3) Coordinates of these regions were transformed from the SST to the individual DSI datasets via the transformation matrix between SST and individual DSI. 4) GFA values were calculated and the mean GFA value within each ROI was calculated for each individual mice. One-way ANOVA was applied to compare GFA values among three groups.Comparing with the control mice, the AD mice without medication showed significantly increased GFA in the hippocampus and thalamus, whereas the AD mice with medication showed no significant difference (Figure 1). In this study, we found that AD mice without medication showed aberrantly increased GFA in the Hp and Th. It has been known that Hp is one of the brain regions that are affected early in the course of AD and is involved in memory and learning function. Thalamus is a region that has close functional connection to the Hp. Qin et al. found that the FA values of APP/PS1 mice were higher than those of the WT controls in several brain regions1. This might be due to the mixture of multiple pathological changes, such as robust amyloid and neuritic plaque deposition, glial hypertrophy, amyloid angiopathy with perivascular space dilation, and swollen neuronal process2,3,4. In our study, the regions with increased GFA values in AD mice corresponded to those regions with increased FA values in a previous study5. Moreover, we observed that GFA in the left Hp and left Th resumed a normal level in the medication group, implying that these two regions seem to undergo a recovery process after two months of treatment.Our findings imply that DSI can be used to monitor the drug effects in AD mice. The neuroprotective drug tested on the AD mice might have positive effects on the Hp and Th. Further works will be conducted to validate the change in GFA by behavioral and histological correlations.