Item

Abstract

Emerging MRI and fMRI methods, most effective at high magnetic field, provide details of human brain white matter structure, cortical architecture and function that are very difficult to include in analysis methods established in the 1990’s—for example SPM and FSL. For essentially pragmatic reasons, these early methods for analysing spatial maps of functional brain activity and assessing structural connectivity incorporated several assumptions that fail to approximate actual neural operations in any living brain. Some of these assumptions were made explicit, but others remain implicit and unexamined. However, the improved data quality now available at 7T, with spatial resolution typically well below the thickness of cortical grey matter, allows the more unrealistic assumptions to be discarded, opening a way forward to far more realistic methods for brain functional analysis. To give specific examples of widely held incorrect assumptions, it cannot be assumed that changes in brain activity are generally spatially smooth [1], nor that every voxel in white matter contains axonal fibres with only one orientation [2]. In this presentation, principles of neural organization and function that should be respected by analysis techniques are listed, basic characteristics of MRI and fMRI signal are described, and some of the new improved analysis methods are outlined. Functional analyses based on in-vivo cortical parcellation [3,4] will have much greater sensitivity and spatial specificity than existing techniques, and will enable much deeper understanding of the co-operative action of neurons across the brain in task accomplishment. Neuroscientific understanding of the relationships between structure, function and connectivity in anatomically distinct brain areas may at last begin to catch up with the achievements of nephrology.