Alzheimer's disease (AD) afflicts 6.5 million individuals in US and 15 million worldwide and elicits a huge economic burden to the society. However, there is no cure for AD and there is an urgent need to identify AD before occurrence of cognitive symptoms/irreversible pathological changes, thus allowing early interventions to delay the onset and progression of the disease. Cerebral vascular dysfunction is among the earliest pathological changes implicated in AD pathogenesis and a potential marker for early AD diagnosis, but the evaluation of the alterations in human cerebral vascular network function is limited by high cost, and low availability and sensitivity. Studies have demonstrated that many AD pathologies also occur in retina and back of eye, including Aβ deposition and microvascular dysfunction. Retinal and cerebral microvasculature share similar structural and physiological properties and can be directly imaged by OCTA. Indeed, vascular dysfunctions in AD retina has been detected in AD patients, including lower capillary density, reduced blood flow and pericyte loss, similar to cerebral vascular changes in AD. However, it remains largely unknown how the ONH and retrolaminar microvascular connecting to brain changes are related to other AD pathological changes because OCTA cannot penetrate back of retina. The goal of this proposal is to use high frequency ultrasound flow image to fill up these knowledge gaps by examining the longitudinal change of ONH and retrolaminar microvasculature in AD mice. There are two aims: Aim1 will examine the temporal changes of vascular function, including blood flow, capillary density, and pericyte loss in different pathological stages of AD mice (5XFAD) by high resolution ultrasound; Aim 2 will determine the relationship between the microvascular alterations and deposition in both brain and back of eye, BBB breakdown and behavior change in AD mice.