Toward layer-specific BOLD fMRI in human cortex at 3T using 3D zoomed-EPI and smallip fast-recovery imaging

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The goal of this project is to develop a sensitive method for "zoomed", or reduced FOV (rFOV), high-resolution functional magnetic resonance imaging (fMRI) at 3T, and apply it to the study of layer-specific activation across the human cortex. Several groups are currently studying the laminar contributions to the observed fMRI signal, with the goal of relating ob- served fMRI signals to underlying neuronal activity. These studies are generally done at high field with mm to sub-mm resolution across layers. In humans, rFOV imaging at 7T has produced the most detailed depictions to date of laminar functional organization. However, the 7T platform presents certain practical and technical challenges, and is not yet widely available to the neuroimaging community. We propose to develop a protocol for rFOV high-resolution 3D functional brain mapping on 3T clinical scanners with enhanced sensitivity and improved spatio-temporal resolution compared to existing rFOV schemes. 3T MRI is now a mature technology and widely available, and the ability to study laminar functional organization at this field strength, and to differentiate between several underlying neuronal processes based on spatio-temporal fMRI signatures identified in high-resolution laminar studies, would be of fundamental importance to the neuroimaging community. The proposed method could ultimately benefit clinical research and practice as well, e.g., by enabling assessment of the layer-specific functional impact of cortical lesions in neurodegenerative diseases. Our approach contains two key ingredients: First, a novel pulse sequence that encodes functional contrast not only in the transverse magnetization produced by the most recent RF excitation pulse, but also in the steady-state longitudinal magnetization. This permits 3D segmented imaging with short TR, which provides high image quality and time-efficient scanning, without loss of functional contrast relative to conventional 2D multislice BOLD. Second, novel 3D tailored RF pulses for rFOV selection that enable fast non-cartesian data readouts with improved temporal resolution compared to standard approaches. We will perform simulation and human volunteer studies to evaluate the proposed sequence, in terms of functional contrast-to-noise ratio (CNR) across cortical layers and ability to resolve layer-specifc activation and BOLD dynamics (e.g., onset time and activation duration).