Presenter: Long Xie
Title: Quantifying Subtle Morphological Changes of the Medial Temporal Lobe Subregions in Early Alzheimer’s Disease
The medial temporal lobe (MTL) is the first cortical region affected by neurofibrillary tangle pathology in Alzheimer’s disease (AD). Quantification of morphological changes in different subregions of the MTL using structural MRI has important utility in tracking disease progression and in clinical trials targeting preclinical or prodromal AD. However, large anatomical variability in the MTL cortex and the confound of the dura (part of the meninges, has similar intensity as gray matter in T1-weighted MRI) hinder our ability in accurately measure the subtle disease-related changes in early AD. In this talk, I will introduce a novel analysis pipeline that is tailored for the quantification of subtle morphological changes in the MTL addressing the above difficulties. In addition, our latest effort in fusing multi-atlas segmentation with deep neural network to improve the MTL subregion segmentation will be discussed. To demonstrate the clinical utility of the proposed method, I will show how we use the extracted volume and regional thickness to detect AD in its prodromal and even preclinical stages as well as to demonstrate their relation to other AD biomarkers of tau and amyloid pathologies.
Bio: Long Xie is a research scientist at the Penn Image Computing and Science Lab (PICSL) at the University of Pennsylvania. Dr. Xie received his doctoral degree in Bioengineering from the University of Pennsylvania. His research is in the field of computational medical image analysis, which involves developing new algorithms for image data analysis and using them to discover how Alzheimer's and related diseases affect the brain. One focus of his work is in developing novel imaging biomarkers using structural and functional MRI for early diagnosis and monitoring of Alzheimer’s disease.
Posted by: Hong Xu