Search

MRI connectomics methods treat the brain as a network and provide new information about its organization, efficiency, and mechanisms of disruption. The most commonly used method of defining network nodes is to register the brain to a standardized anatomical atlas based on the Brodmann areas. This approach is limited by inter-subject variability and can be especially problematic in the context of brain maturation or neuroplasticity (cerebral reorganization after brain damage). In this study, we combined different image processing and network theory methods and created a novel approach that enables atlas-free construction and connection-wise comparison of diffusion MRI-based brain networks. We illustrated the proposed approach in three age groups: neonates, 6-month-old infants, and adults. First, we explored a data-driven method of determining the optimal number of equal-area nodes based on the assumption that all cortical areas of the brain are connected and, thus, no part of the brain is structurally isolated. Second, to enable a connection-wise comparison, alignment to a "reference brain" was performed in the network domain within each group using a matrix alignment algorithm with simulated annealing. The correlation coefficients after pair-wise network alignment ranged from 0.6102 to 0.6673. To test the method's reproducibility, one subject from the 6-month-old group and one from the adult group were scanned twice, resulting in correlation coefficients of 0.7443 and 0.7037, respectively. While being less than 1 due to parcellation and noise, statistically, these values were significantly higher than inter-subject values. Rotation of the parcellation largely explained the variability. Through the abstraction from anatomy, the developed framework allows for a fully network-driven analysis of structural MRI connectomes and can be applied to subjects at any stage of development and with substantial differences in cortical anatomy.

Neonatal encephalopathy represents a heterogeneous group of conditions associated with life-long developmental disabilities and neurological deficits. Clinical measures and current anatomic brain imaging remain inadequate predictors of outcome in children with neonatal encephalopathy. Some studies have suggested that brain development and, therefore, brain connectivity may be altered in the subgroup of patients who subsequently go on to develop clinically significant neurological abnormalities. Large-scale structural brain connectivity networks constructed using diffusion tractography have been posited to reflect organizational differences in white matter architecture at the mesoscale, and thus offer a unique tool for characterizing brain development in patients with neonatal encephalopathy. In this manuscript we use diffusion tractography to construct structural networks for a cohort of patients with neonatal encephalopathy. We systematically map these networks to a high-dimensional space and then apply standard machine learning algorithms to predict neurological outcome in the cohort. Using nested cross-validation we demonstrate high prediction accuracy that is both statistically significant and robust over a broad range of thresholds. Our algorithm offers a novel tool to evaluate neonates at risk for developing neurological deficit. The described approach can be applied to any brain pathology that affects structural connectivity.

The corpus callosum is hypothesized to play a fundamental role in integrating information and mediating complex behaviors. Here, we demonstrate that lack of normal callosal development can lead to deficits in functional connectivity that are related to impairments in specific cognitive domains. We examined resting-state functional connectivity in individuals with agenesis of the corpus callosum (AgCC) and matched controls using magnetoencephalographic imaging (MEG-I) of coherence in the alpha (8-12 Hz), beta (12-30 Hz) and gamma (30-55 Hz) bands. Global connectivity (GC) was defined as synchronization between a region and the rest of the brain. In AgCC individuals, alpha band GC was significantly reduced in the dorsolateral pre-frontal (DLPFC), posterior parietal (PPC) and parieto-occipital cortices (PO). No significant differences in GC were seen in either the beta or gamma bands. We also explored the hypothesis that, in AgCC, this regional reduction in functional connectivity is explained primarily by a specific reduction in interhemispheric connectivity. However, our data suggest that reduced connectivity in these regions is driven by faulty coupling in both inter- and intrahemispheric connectivity. We also assessed whether the degree of connectivity correlated with behavioral performance, focusing on cognitive measures known to be impaired in AgCC individuals. Neuropsychological measures of verbal processing speed were significantly correlated with resting-state functional connectivity of the left medial and superior temporal lobe in AgCC participants. Connectivity of DLPFC correlated strongly with performance on the Tower of London in the AgCC cohort. These findings indicate that the abnormal callosal development produces salient but selective (alpha band only) resting-state functional connectivity disruptions that correlate with cognitive impairment. Understanding the relationship between impoverished functional connectivity and cognition is a key step in identifying the neural mechanisms of language and executive dysfunction in common neurodevelopmental and psychiatric disorders where disruptions of callosal development are consistently identified.

Defining the structural and functional connectivity of the human brain (the human "connectome") is a basic challenge in neuroscience. Recently, techniques for noninvasively characterizing structural connectivity networks in the adult brain have been developed using diffusion and high-resolution anatomic MRI. The purpose of this study was to establish a framework for assessing structural connectivity in the newborn brain at any stage of development and to show how network properties can be derived in a clinical cohort of six-month old infants sustaining perinatal hypoxic ischemic encephalopathy (HIE). Two different anatomically unconstrained parcellation schemes were proposed and the resulting network metrics were correlated with neurological outcome at 6 months. Elimination and correction of unreliable data, automated parcellation of the cortical surface, and assembling the large-scale baby connectome allowed an unbiased study of the network properties of the newborn brain using graph theoretic analysis. In the application to infants with HIE, a trend to declining brain network integration and segregation was observed with increasing neuromotor deficit scores.