Alzheimer’s disease (AD) is a devastating neurodegenerative disorder that is typically diagnosed at late stages based on cognitive decline. The presence of beta amyloid (Ab) plaques, which has been the subject of study for over a century, emphasizes abnormal Aβ protein production as an upstream causative factor of AD. Microglial activation and oxidative stress have also been linked to the formation of amyloid plaques found in AD. The ability to non-invasively detect AD plaques with ultra high field (≥ 7 Tesla) magnetic resonance imaging (MRI) has been demonstrated in post mortem human AD specimens, as well as in vivo in transgenic murine models (1, 2). However, this feat has not been accomplished in larger animal models. We previously demonstrated that rabbits fed cholesterol-enriched diets develop iron-rich Ab-enriched plaques in their brains and that we could detect these plaques in ex vivo brain specimens using highly specialized low-volume RF hardware, iron-sensitive pulse sequences, and a clinical-field strength (3 Tesla (T)) MR scanner (3). Our aim in the present work was to develop an in vivo MRI protocol at a clinical strength capable of non-invasively visualizing plaque burden in this rabbit model of AD and perform further evaluation of brain specimens for other hallmarks of human AD such as microglial activation.Rabbits were fed a cholesterol-rich (0.125% to 0.25%) diet to maintain a serum cholesterol level of ~400 mg/dl (n=8) or a normal chow diet (n=5) for 24 months. To prevent liver failure cholesterol was removed from all diets for an additional 8-14 months. In vivo MRI was performed using a two-channel phased array surface RF coil on a 3T scanner. Three-dimensional (3D) fast imaging employing steady state acquisition (FIESTA) imaging was performed in the axial plane with a resolution 200x200x200 μm3, TR/TE 20/10 msec, FA=20, and a scan time of 83 minutes. Susceptibility-weighted post-processing of FIESTA images (SWI-FIESTA) was performed in Matlab by using a k-space filter to generate a phase map that enhances regions of high frequency phase change in the image. The phase map was converted to a phase mask and then multiplied into the magnitude image 4 times. In vivo 3D susceptibility weighted imaging with multi-echo acquisition (SWAN) using a 3D multi-echo gradient echo sequence to enhance the T2* effect was also performed in the axial plane with a resolution 200x200x200 µm3, TR/TE 83.1/29.9 ms and a scan time of 86 minutes. A trained radiologist blind to diet conditions manually counted plaque associated signal void/hypointensity in different brain regions based on criteria established in our previous ex vivo studies. Following MRI, microglia immunostaining, myeloperoxidase immunostaining, Ab immunostaining and Prussian blue iron staining were performed on brain sections.In vivo MRI revealed signal voids/hypointensities in the brains of the cholesterol-fed (Fig. 1A/B) but not control rabbits (Fig. 1C/D). Image analysis confirmed significantly more voids in the cortex, sub-cortex, and hippocampus of cholesterol-fed rabbits (p<0.01; Fig. 1E/F). SWI processing of FIESTA images significantly improved the detectability of plaques in the cortex and sub-cortical regions (data not shown). Subsequent Ab and iron staining showed that voids in MR images corresponded to iron-laden Ab-positive plaques in matched brain sections (Fig. 2). Further histological analysis showed that microglia were found in close proximity to these plaques and that these microglia were activated with presence of positive myeloperoxidase immunoreactivity (data not shown).In this study we show that a low level of cholesterol diet in rabbits induces cerebral Aβ-positive iron-enriched AD plaques with associated microglial activation; key characteristics also seen in human AD but often lacking in transgenic mouse models. In conjunction with this, we have developed MRI protocols capable of non-invasively visualizing iron-rich plaques in vivo using a relatively higher volume RF coil than previously used in our ex vivo studies. This technology will allow us to study of the natural history of plaque formation in this model, as well as monitoring the efficacy of novel interventions in individual animals over time. Our ability to detect AD plaques in vivo using clinical field strength MRI with conventional hardware is an important step towards clinical translation of direct AD plaque detection in humans.