Performing label-fusion-based segmentation using multiple automatically generated templates.

Classically, model-based segmentation procedures match magnetic resonance imaging (MRI) volumes to an expertly labeled atlas using nonlinear registration. The accuracy of these techniques are limited due to atlas biases, misregistration, and resampling error. Multi-atlas-based approaches are used as a remedy and involve matching each subject to a number of manually labeled templates. This approach yields numerous independent segmentations that are fused using a voxel-by-voxel label-voting procedure. In this article, we demonstrate how the multi-atlas approach can be extended to work with input atlases that are unique and extremely time consuming to construct by generating a library of multiple automatically generated templates of different brains (MAGeT Brain). We demonstrate the efficacy of our method for the mouse and human using two different nonlinear registration algorithms (ANIMAL and ANTs). The input atlases consist a high-resolution mouse brain atlas and an atlas of the human basal ganglia and thalamus derived from serial histological data. MAGeT Brain segmentation improves the identification of the mouse anterior commissure (mean Dice Kappa values (κ = 0.801), but may be encountering a ceiling effect for hippocampal segmentations. Applying MAGeT Brain to human subcortical structures improves segmentation accuracy for all structures compared to regular model-based techniques (κ = 0.845, 0.752, and 0.861 for the striatum, globus pallidus, and thalamus, respectively). Experiments performed with three manually derived input templates suggest that MAGeT Brain can approach or exceed the accuracy of multi-atlas label-fusion segmentation (κ = 0.894, 0.815, and 0.895 for the striatum, globus pallidus, and thalamus, respectively).

Sex-specific associations between subcortical morphometry in childhood and adult alcohol consumption: A 17-year follow-up study.

Men and women tend to differ in the age of first alcohol consumption, transition into disordered drinking, and the prevalence of alcohol use disorder. Here, we use a unique longitudinal dataset to test for potentially predispositonal sex-biases in brain organization prior to initial alcohol exposure. Our study combines measures of subcortical morphometry gathered in alcohol naive individuals during childhood (mean age: 9.43 years, SD = 2.06) with self-report measures of alcohol use in the same individuals an average of 17 years later (N = 81, 46 males, 35 females). We observe that pediatric amygdala and hippocampus volume both show sex-biased relationships with adult drinking. Specifically, females show a stronger association between subcortical volumetric reductions in childhood and peak drinking in adulthood as compared to males. Detailed analysis of subcortical shape localizes these effects to the rostro-medial hippocampus and basolateral amygdala subnuclei. In contrast, we did not observe sex-specific associations between striatal anatomy and peak alcohol consumption. These results are consistent with a model in which organization of the amygdala and hippocampus in childhood is more relevant for subsequent patterns of peak alcohol use in females as compared to males. Differential neuroanatomical precursors of alcohol use in males and females could provide a potential developmental basis for well recognized sex-differences in alcohol use behaviors.. Thus, our findings not only indicate that brain correlates of human alcohol consumption are manifest long before alcohol initiation, but that some of these correlates are not equivalent between males and females.

The Evolution of Cerebral Language Localization: Historical Analysis and Current Trends.

Language localization has been an evolving concept over the past 150 years, with the emergence of several important yet conflicting ideologies. The classical theory, starting from the phrenologic work of Gall to the identification of specific regions of language function by Broca, Wernicke, and others, proposed that discrete subcomponents of language were organized into separate anatomic structural regions. The holism theory was postulated in an attempt to disclose that language function was instead attributed to a larger region of the cortex, in which cerebral regions may have the capability of assuming the function of damaged areas. However, this theory was largely abandoned in favor of discrete structural localizationist viewpoints. The subsequent cortical stimulatory work of Penfield led to the development of maps of localization, assigning an eloquent designation to specific regions. The expanding knowledge of cortical and subcortical anatomy allowed for the development of anatomically and functionally integrative language models. In particular, the dual stream model revisited the concept of regional interconnectivity and expanded the concept of eloquence. Advancements in cortical-subcortical stimulation, neurophysiologic monitoring, magnetic resonance diffusion tensor imaging/functional magnetic resonance imaging, awake neurosurgical technique, and knowledge gained by white matter tract anatomy and the Human Connectome Project, shed new light on the dynamic interconnectivity of the cerebrum. New studies are progressively opening doors to this paradigm, showing the dynamic and interdependent nature of language function. In this review, the evolution of language toward the evolving paradigm of dynamic language function and interconnectivity and its impact on shaping the neurosurgical paradigm are outlined.