Unraveling the mesoscopic functional organization of the human visual cortex using high-field MRI

点击翻译按钮获取中文摘要

The uniformity of the mammalian cortex has led to the proposition that there exist elementary cortical units of operation, consisting of several hundred or thousand neurons that are repeated within and across cortical areas (Lorente de No', 1938). Cortical columns and layers of the neocortex are prominent examples of such structurally and functionally specialized units. These fine-scale anatomical constructs appear throughout cortical areas. Based on numerous animal experiments, the functional properties are similar for neurons within a column but are known to vary between columns. It can, therefore, be argued that the optimal spatial scale for studying the relationship between brain function and behavior is that of cortical columns (and layers, for similar reasons). The responses of these fine-scale structures to visual stimuli as well as their spontaneous activity can now be probed in the human brain by combining high-resolution functional MRI (fMRI) at ultra-high (7 Tesla or higher) magnetic field and analysis methods specifically tailored for this task. The long term goal of my research program is to unravel the functional organization, the function, and the interactions of cortical columns and layers in lower visual areas of the human cerebral cortex, in order to understand their mechanistic roles in support of perception. Our first objective is to develop a novel method for detecting and correcting small head movements in high-resolution fMRI time-series of a small brain volume. In our second objective, we will compare the spatial specificity of cerebral blood volume-based fMRI and Gradient and Spin-Echo (GRASE) blood oxygenation level-dependent (BOLD) fMRI in the human visual cortex. Our third objective is to compute the spatial point-spread function of cerebral blood volume-based fMRI and GRASE BOLD fMRI in the human visual cortex. Our fourth and fifth objectives are to simultaneously image, and quantitatively characterize and model the complete maps of ocular dominance columns and orientation columns, respectively, from the entire primary visual area of human subjects. To realize these objectives, we will image human subjects using the recently installed Siemens 7 Tesla Terra scanner of the Brain Imaging Centre of the Montreal Neurological Institute and McGill University. We will apply model-based analysis as well as data-driven machine learning techniques. We expect that the findings from addressing objectives 1-3 will significantly improve high-resolution fMRI at the scale of cortical columns and layers. They can provide the basis for using fMRI and decoding techniques for discovering fine-scale functional organizations in regions of the brain where fine cortical maps have not been revealed, such as higher-order association areas. The results of addressing aims 4-5 will set the stage for investigating the neuronal mechanisms of human perception of shape, stereopsis and binocular rivalry at the scale of cortical columns.