Greg Harris

Structural MR image processing using the brains2 toolbox

Medical imaging has opened a new door into biomedical research. In order to study various diseases of the brain and detect their impact on brain structure, robust and user friendly image processing packages are required. These packages must be multi-faceted to distinguish variations in size, shape, volume, and the ability to detect longitudinal changes over the course of an illness. This paper describes the BRAINS2 image processing package, which contains both manual and automated tools for structural identification, methods for tissue classification and cortical surface generation. These features are described in detail, as well as the reliability of these procedures.

Abnormalities in midline attentional circuitry in schizophrenia: evidence from magnetic resonance and positron emission tomography.

The syndrome of schizophrenia presents with a complex array of symptoms that are difficult to explain at the neural level. Data collected using magnetic resonance (MR) and positron emission tomography (PET) suggest that this complex array could occur as a consequence of misconnections and mismatches in midline circuitry that is reticular-thalamic-cingulate-cortical. MR studies have shown a variety of abnormalities, including callosal agenesis, cavum septi pellucidi, decreased thalamic size, decreased frontal size, and changes in signal intensity in white matter tracts between the thalamus and the frontal cortex. PET studies using a dichotic listening paradigm suggest that patients suffering from schizophrenia have brain blood flow abnormalities consistent with a difficulty in focusing or shifting attention, which may reflect the functional substrate of the anatomic abnormalities.

Image registration issues in the analysis of multiple injection 15O H2O PET studies: BRAINFIT

Spatial registration of functional and structural data is a vital first step in the regional analysis of images obtained from functional imaging techniques. We present a set of procedures that produce accurate, objective, and robust registrations between PET and MR images that are highly automated and user-friendly. These techniques have been shown to be both reliable and valid. Because the method is automated and therefore efficient, each injection of the PET experiment can be individually fit to overcome errors introduced because of subject movement between PET injections.

Brain structures defined by an artificial neural network

In order to study brain changes associated with various psychiatric and neurological disorders, comparisons need to be made to normal controls since these diseases often have very subtle differences (i.e. small effect sizes) in brain development and maturation. In order to find these differences often fairly large samples need to be studied. Defining several regions on such sample sizes can pose a rather daunting task which requires a substantial time investment and is subject to rater drift and reliability issues as several traces may trace a structure over an extended period of time. Automated procedures which are reliable ease the burden of such studies. The may include atlas based methods [l], non-linear warping [2-31, or artificial intelligence based methods [4]. Developing such methods allow for the rapid study of an ever increasing number of structures which can he defined via magnetic resonance imaging.

In vivo measurement of the thalamic subregions using contrast-enhanced MRI

1&duction: The thalamus is a major relay nucleus for much of the information processed in the brain; as such the thalamus is of interest to many areas of brain research including, normal development, various pathologies, and aging. The thalamus is a diverse structure, composed of numerous subnuclei serving distinct regions of the cortex. Examination of these subnuclei is necessary for a meaningful examination of deficit syndromes as well as normal functioning of specific cortical regions. The current study examines the major nuclear groups divided using the intramedullary lamina (the anterior, medial, lateral and geniculate nuclei as well as the pulvinar).