Recent findings from clinical trials of pharmacological interventions in Alzheimer’s Disease (AD) such as lecanemab have shown a significant slowing of clinical decline in early-stage patients. However, potentially serious side effects such as amyloid-related imaging abnormalities (ARIA) caused by brain swelling and bleeding highlight the importance of detecting AD early in its course so that treatment can (1) be applied when it is most effective, prior to widespread neuronal loss, and (2) avoid potentially harmful side effects in patients that would not benefit from treatment. Thus, for effective treatment of AD it is critical that we develop sensitive and specific surrogate biomarkers that accurately distinguish subjects with AD from other populations as early as possible in the course of the disease, as well as quantify the impacts of AD on brain anatomy. Unfortunately, this is a difficult undertaking using images of an individual at a single time point due to the enormous variability in normal human neuroanatomy, which can obscure disease processes. Compared to a cross-sectional approach, a longitudinal design can provide more sensitivity and specificity for examining subtle associations by reducing the confounding effect of between-subject variability. Moreover, a serial assessment can be the only way to unambiguously characterize the effect of interest in a randomized experiment, such as a drug trial. Finally, longitudinal studies can provide unique insights into the temporal dynamics of the biological processes underlying AD disease progression. Taking full advantage of a longitudinal design requires the optimization of the computational tools that perform image processing and hypothesis testing. In this project, we propose to design, develop, and distribute Deep Learning-based, intrinsically longitudinal image processing tools that incorporate models summarizing the neuroanatomical effects of AD progression. Specifically, we will develop a model of the temporal effects of AD on brain anatomy allowing us to classify AD patients from controls and subjects with other neurodegenerative diseases, then embed that model into a longitudinal segmentation framework, resulting in a set of open-source and publicly distributed algorithms that are specifically designed to be optimal for diagnosing and quantifying the effects of AD. These tools will then be provided to the ADNI community and applied to ADNI data, as we have done in the past.