DESCRIPTION (Investigator's Abstract): This project is designed to assess the utility of serial magnetic resonance (MR)-based volumetric measurements of the anterior temporal lobe (ATL) and hippocampal formation (HF) in aiding in the clinical diagnosis of dementia of the Alzheimer's type (DAT). This project will use MR volumes to provide predictive information on patients with a mild cognitive impairment as to which patients will progress to DAT and which will remain cognitively stable. Our previous work has shown that this MR-based volume measurement technique is useful in differentiating between DAT and normal controls; however, a single scan on patients in the mild cognitive impairment group is less definitive. We will draw upon previously funded projects to provide appropriately characterized patients for this project. The Mayo Clinic Alzheimer's Disease Patient Registry and the Alzheimer's Disease Center will provide the mechanism for recruitment, evaluation, and longitudinal assessment of patients and controls. An additional ongoing project funded by the NINDS will provide initial baseline MR scans on the patients and controls. Thus, the present application proposes to perform serial MR scans on three groups of individuals, DAT, mild cognitive impairment, and normal controls, to determine if volumetric changes in the ATL and HF will improve discrimination among the groups. We will be particularly interested in determining if volumetric changes of these structures help to determine which patients in the mild cognitive impairment group will ultimately progress to DAT and which patients will remain clinically stable. We will also correlate changes in the volumes of the ATL and HF with changes in clinical performance and cognitive function (particularly memory) over the course of the study. This proposal represents a unique opportunity at this point in time to maximally utilize existing resources to address important issues regarding the role of MR volume try in improving the early diagnosis of DAT.