Grant: Discovery Grants Program - Individual

Magnetic Resonance Imaging (MRI) is a versatile tool that allows us to probe a large array of brain physiological properties. However, most quantitative techniques to measure brain physiology are onerous, typically requiring difficult to obtain custom and often lengthy sequences, specialized equipment or physiological challenges such as gas inhalation. Therefore, most large-scale datasets do not include such quantitative methods. Recent developments have however made some of these methods more accessible, including a novel gas-free method for measuring cerebrovascular reactivity (CVR) using resting state blood oxygen level dependent (BOLD) signal and quantitative susceptibility mapping (QSM) to measure oxygen extraction fraction (OEF). Here, we assess the validity of these methods as compared to more established quantitative techniques, including hypercapnia CVR and calibrated functional MRI (fMRI) in older adults, and assess their sensitivity to detect the effects of lifestyle. Our overarching objective is to create an accessible framework to measure CVR and OEF in existing publicly available large datasets and assess their ability to measure known physiological variability. Our specific objectives are: Objective 1 (O1). O1.1 Assess the validity of a novel resting state CVR measure and its sensitivity to detect the effects of fitness; O1.2 Assess its sensitivity to detect sex and PA dose (O1.2); Objective 2 (O2). Assess the sensitivity of resting state CVR imaging to the effects of diet and sleep; Objective 3 (O3). O3.1 Improve QSM acquisition and venous segmentation; and O3.2 assess the validity of QSM in older populations and its sensitivity to detect the effects of sex and fitness; Methodology: To address these objectives, we will use a combination of three datasets. The ACTION dataset includes 1-year longitudinal data in 150 healthy older adults, with QSM, resting state BOLD, hypercapnia and calibrated fMRI, as well as fitness measures. This dataset will allow us to address objectives O1.1 and 3.2. The Human Connectome Project - Aging includes resting state BOLD and PA information, allowing us to assess O1.2. The UK Biobank includes resting state BOLD, 7-day accelerometry and diet questionnaires, allowing us to address O2. All analyses will be performed in sex-disaggregated data and take into account the effect of race when available. Finally, we will assess whether non-cartesian trajectory readouts and a deep learning venous segmentation approach can be used to improve QSM data. Sensitivity to gender will be assessed when available. Impact: Our ability to detect subtle changes related to lifestyle in the brain is dependent on our access to imaging methods which are both sensitive and quantitative, to limit biases. Validation of quantitative methods which are less onerous to acquire would open to door to new studies into healthy aging and diseases.