DEFINING FUNCTIONAL AREAS IN THE HUMAN BRAIN USING RESTING FUNCTIONAL CONNECTIVIT

DESCRIPTION (provided by applicant): The cerebral cortex is anatomically organized at many physical scales. This organization extends from the level of a neuron and its connections, collections of neurons into columns, collections of columns into functional areas or maps, and collections of areas into systems or networks. Current functional magnetic resonance imaging (fMRI) techniques in humans focuses above the level of cells and columns, most often at the scale of functional areas and systems. Except in rare circumstances, e.g. topographically-organized sensory regions, it is difficult to determine areal boundaries in the human brain using fMRI. The ability to non- invasively delineate functional areas in greater extents of cortex would allow both within-subject definition of separate areas of cortex, and allow for more appropriate comparison of areas across subjects, enhancing the precision of many types of both functional and structural studies, including those investigating special populations such as infants/children and patients. Preliminary measures of resting correlated activity, so-called resting state functional connectivity (fcMRI), show strong localized differences in correlation strength across expanses of cortex. This observation provides hope that, in similar fashion to connectional anatomy in non-human primates, fcMRI measures could aid in the definition of functional area boundaries in the human brain. Aim 1: To develop methods to define functional area boundaries using resting state fcMRI. We will utilize the relatively new metric of resting state fcMRI to define boundaries between putative functional areas based on abrupt changes in the correlation profiles of adjacent points in fcMRI data. Various image processing strategies (e.g., edge detection and image segmentation algorithms) will be used to produce maps of fcMRI-derived boundaries that circumscribe putative functional areas in the cortex. Aim 2: To validate methods using multiple converging tests. After putative fcMRI boundary maps are constructed, their reliability and validity will be tested using several converging approaches: 1) repeated within-subject examinations to test the reliability of fcMRI-derived boundaries; 2) comparison of functionally-defined (fMRI) boundaries in topographically organized visual cortical regions with fcMRI-defined borders; 3) Comparison of reliable functional activations (e.g. eye- movement and error-related activity) with fcMRI-defined regions to determine whether functional activations respect fcMRI-derived boundaries; 4) examination of multiple subjects to test the generalizability of these maps. PUBLIC HEALTH RELEVANCE This grant proposes a series of methods development and validations for using a specific imaging measure, resting-state functional connectivity magnetic resonance imaging, to parcellate the cortex of the brain into individual functional regions in individual subjects. Successful developments would allow researchers to move beyond current macrostructural, stereotactic and combined methods to provide more accurate locations of functional areas. This would advance neuroimaging studies generally, but particularly in situations where individual or group variability would be high, such as development studies, and studies of neuropsychiatric and neurological disease including schizophrenia, depression, movement disorders, and stroke.

DESCRIPTION (provided by applicant): The cerebral cortex is anatomically organized at many physical scales. This organization extends from the level of a neuron and its connections, collections of neurons into columns, collections of columns into functional areas or maps, and collections of areas into systems or networks. Current functional magnetic resonance imaging (fMRI) techniques in humans focuses above the level of cells and columns, most often at the scale of functional areas and systems. Except in rare circumstances, e.g. topographically-organized sensory regions, it is difficult to determine areal boundaries in the human brain using fMRI. The ability to non- invasively delineate functional areas in greater extents of cortex would allow both within-subject definition of separate areas of cortex, and allow for more appropriate comparison of areas across subjects, enhancing the precision of many types of both functional and structural studies, including those investigating special populations such as infants/children and patients. Preliminary measures of resting correlated activity, so-called resting state functional connectivity (fcMRI), show strong localized differences in correlation strength across expanses of cortex. This observation provides hope that, in similar fashion to connectional anatomy in non-human primates, fcMRI measures could aid in the definition of functional area boundaries in the human brain. Aim 1: To develop methods to define functional area boundaries using resting state fcMRI. We will utilize the relatively new metric of resting state fcMRI to define boundaries between putative functional areas based on abrupt changes in the correlation profiles of adjacent points in fcMRI data. Various image processing strategies (e.g., edge detection and image segmentation algorithms) will be used to produce maps of fcMRI-derived boundaries that circumscribe putative functional areas in the cortex. Aim 2: To validate methods using multiple converging tests. After putative fcMRI boundary maps are constructed, their reliability and validity will be tested using several converging approaches: 1) repeated within-subject examinations to test the reliability of fcMRI-derived boundaries; 2) comparison of functionally-defined (fMRI) boundaries in topographically organized visual cortical regions with fcMRI-defined borders; 3) Comparison of reliable functional activations (e.g. eye- movement and error-related activity) with fcMRI-defined regions to determine whether functional activations respect fcMRI-derived boundaries; 4) examination of multiple subjects to test the generalizability of these maps. PUBLIC HEALTH RELEVANCE This grant proposes a series of methods development and validations for using a specific imaging measure, resting-state functional connectivity magnetic resonance imaging, to parcellate the cortex of the brain into individual functional regions in individual subjects. Successful developments would allow researchers to move beyond current macrostructural, stereotactic and combined methods to provide more accurate locations of functional areas. This would advance neuroimaging studies generally, but particularly in situations where individual or group variability would be high, such as development studies, and studies of neuropsychiatric and neurological disease including schizophrenia, depression, movement disorders, and stroke.