Multiband ASL for Alzheimer's Disease

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT Alzheimer’s disease (AD) is the most common cause of dementia with enormous healthcare burden. Identification of preclinical disease is critical for the development of therapy, as significant neuronal death has already occurred by the time of symptom onset. Arterial spin labeled (ASL) perfusion MRI is an appealing approach for measuring perfusion in dementia by utilizing magnetically labeled arterial blood water as an endogenous tracer. We and others have applied ASL to study AD and mild cognitive impairment (MCI). Characteristic patterns of cerebral hypoperfusion in temporoparietal association cortices, posterior cingulate cortex (PCC), precuneus and frontal cortex were detected using ASL in AD patients, and to a lesser extent, in MCI populations. However, there remain several challenges for making ASL an impactful tool in studying AD and other neurodegenerative disorders, including: 1) Existing ASL techniques generally have a coarse spatial resolution of ~4x4x4mm3, making it difficult to decompose structural and functional components of neurodegenerative effects due to partial volume effects; 2) The recommended implementation of pseudo- continuous ASL (pCASL) with segmented 3D acquisition is susceptible to (inter-segment) head motion that is frequently present in aged populations; and 3) The relatively long duration of segmented 3D acquisition generally allows a single post-labeling delay (PLD) scan, which is susceptible to age dependent variations in arterial transit time, affecting the accuracy of perfusion quantification. The goal of the parent R01 project (EB028297 “Multiband ASL for Neurodevelopment Study”) was to develop and evaluate cutting-edge multiband (MB) pCASL protocols that are able to offer a high spatial resolution of isotropic 2mm or higher, resistance to head motion and multi- delay capability for accurate perfusion quantification in pediatric populations. During this project, we will apply cutting edge high-resolution 2D and 3D MB pCASL techniques and deep-learning (DL) denoising algorithms in 3 groups of mild AD, amnestic MCI and age matched control subjects, and compare the results with those by standard pCASL methods. We hypothesize that the developed MB pCASL protocols and DL algorithms are more sensitive than standard pCASL techniques for detecting perfusion differences between mild AD, MCI and control subjects. The successful completion of this project will lead to robust high resolution multi-delay MB pCASL protocols with associated DL denoising algorithms which may serve as biomarkers for AD and MCI.