Tracking brain structural trajectories for early detection and prognosis of Alzheimer's Disease - Project Summary Tengfei Li, Ph.D, is a statistician committed to developing advanced statistical methods to better understand the human brain function and structure, the aging process, cognition and brain disorders using multi-modal neuroimaging and genetic data. The research he proposes, entitled “Tracking brain structural trajectories for early detection and prognosis of Alzheimer's Disease”, will unify multi-center multi-modal datasets using cutting-edge statistical and machine learning techniques. The result will be a comprehensive set of life-span brain structural reference standards and a centile-based explainable artificial intelligence (AI) system, which facilitate timely and precise detection of brain structural changes in the early stages of Alzheimer's Disease (AD) and AD related dementias (ADRD) and aid in the diagnosis and prognosis of the early stages of AD. Candidate: Dr. Li is a Research Assistant Professor in the department of Radiology, University of North Carolina at Chapel Hill. He completed his Ph. D in probability and mathematical statistics prior to his postdoctoral fellowship in biostatistics at the University of Texas MD Anderson Cancer Center. Dr. Li's previous work, which has been published in high-impact journals such as Science and Nature Genetics, focuses on imaging-genetic analysis and big data integration. The proposed career development plan will build upon his previous training with four training goals to enhance his trajectory toward becoming an independent investigator: 1) Build a solid understanding of the neurodevelopmental basis of AD/ADRD; 2) Develop novel methodologies and strengthen in-depth understanding of AI in neuroimaging and AD/ ADRD analyses; 3) Develop professional skills to analyze multicenter big data; and 4) Establish leadership, expand collaborative networks, and translate diagnostic tools into clinical practice. Dr. Li and his primary mentor, Hongtu Zhu, Ph.D, have assembled a strong team of co-mentors to guide Dr. Li through the proposed training and research activities. Research: Despite an increased availability of neuroimaging datasets, the analysis of brain structure change by AD is limited by narrow age ranges, small sample sizes, and a lack of critical covariates. Besides, there is no reference standard against which to anchor individual differences in brain morphology from multisite images that may or may not align with growth charts. Dr. Li proposes to overcome these limitations by establishing life-span brain structural developmental reference standards that can be used to identify abnormal brain aging/developmental patterns in the early AD phases, including cortical/subcortical structures and white matter tracts. In Aim 1, he will develop a Functional Harmonization Regression Modeling (FHRM) framework to integrate multi-site data; in Aim 2, he will build a brain Local Growth Curve Modeling (LGCM) system to build the normal brain structural aging trajectories and track their variations associated with early phases or high-risk AD groups; and in Aim 3, he will develop a computationally efficient, easily interpretable and state-of-the-art Centile-based Explainable AI (CEAI) system in delineating the preclinical AD, MCI and AD transitions. This award will provide Dr. Li with the training and research needed to be successful for future, high- impact multi-center multimodal studies and behavioral/cognitive performance related to brain aging and AD. This will significantly contribute to the field and support Dr. Li's transition towards scientific independence.
