Eileen Luders

YNIMG Journal 2022 Journal Article

The neuroanatomy of pregnancy and postpartum

• Eileen Luders
• Florian Kurth
• Inger Sundström Poromaa

Pregnancy and giving birth are exceptional states in a woman's life for many reasons. While the effects of pregnancy and childbirth on the female body are obvious, less is known about their impact on the female brain, especially in humans. The scientific literature is still sparse but we have identified 12 longitudinal neuroimaging studies conducted in women whose brains were scanned before pregnancy, during pregnancy, and/or after giving birth. This review summarizes and discusses the reported neuroanatomical changes during pregnancy, postpartum, and beyond. Some studies suggest that pregnancy is mainly associated with tissue decreases, and a few studies suggest that this tissue loss is mostly permanent. In contrast, the majority of studies seems to indicate that the postpartum period is accompanied by substantial tissue increases throughout the entire brain. Future research is clearly warranted to replicate and extend the current findings, while addressing various limitations and shortcomings of existing studies.

YNIMG Journal 2017 Journal Article

The impact of aging on subregions of the hippocampal complex in healthy adults

• Florian Kurth
• Nicolas Cherbuin
• Eileen Luders

The hippocampal complex, an anatomical composite of several subregions, is known to decrease in size with increasing age. However, studies investigating which subregions are particularly prone to age-related tissue loss revealed conflicting findings. Possible reasons for such inconsistencies may reflect differences between studies in terms of the cohorts examined or techniques applied to define and measure hippocampal subregions. In the present study, we enhanced conventional MR-based information with microscopically defined cytoarchitectonic probabilities to investigate aging effects on the hippocampal complex in a carefully selected sample of 96 healthy subjects (48 males/48 females) aged 18–69 years. We observed significant negative correlations between age and volumes of the cornu ammonis, fascia dentata, subiculum, and hippocampal-amygdaloid transition area, but not the entorhinal cortex. The estimated age-related annual atrophy rates were most pronounced in the left and right subiculum with −0. 23% and −0. 22%, respectively. These findings suggest age-related atrophy of the hippocampal complex overall, but with differential effects in its subregions. If confirmed in future studies, such region-specific information may prove useful for the assessment of diseases and disorders known to modulate age-related hippocampal volume loss.

YNIMG Journal 2016 Journal Article

Estimating brain age using high-resolution pattern recognition: Younger brains in long-term meditation practitioners

• Eileen Luders
• Nicolas Cherbuin
• Christian Gaser

Normal aging is known to be accompanied by loss of brain substance. The present study was designed to examine whether the practice of meditation is associated with a reduced brain age. Specific focus was directed at age fifty and beyond, as mid-life is a time when aging processes are known to become more prominent. We applied a recently developed machine learning algorithm trained to identify anatomical correlates of age in the brain translating those into one single score: the BrainAGE index (in years). Using this validated approach based on high-dimensional pattern recognition, we re-analyzed a large sample of 50 long-term meditators and 50 control subjects estimating and comparing their brain ages. We observed that, at age fifty, brains of meditators were estimated to be 7. 5years younger than those of controls. In addition, we examined if the brain age estimates change with increasing age. While brain age estimates varied only little in controls, significant changes were detected in meditators: for every additional year over fifty, meditators' brains were estimated to be an additional 1month and 22days younger than their chronological age. Altogether, these findings seem to suggest that meditation is beneficial for brain preservation, effectively protecting against age-related atrophy with a consistently slower rate of brain aging throughout life.

YNICL Journal 2014 Journal Article

Neuroprotective effects of testosterone treatment in men with multiple sclerosis

• Florian Kurth
• Eileen Luders
• Nancy L. Sicotte
• Christian Gaser
• Barbara S. Giesser
• Ronald S. Swerdloff
• Michael J. Montag
• Rhonda R. Voskuhl

Multiple sclerosis (MS) is an inflammatory and neurodegenerative disease of the central nervous system. While current medication reduces relapses and inflammatory activity, it has only a modest effect on long-term disability and gray matter atrophy. Here, we have characterized the potential neuroprotective effects of testosterone on cerebral gray matter in a pilot clinical trial. Ten men with relapsing-remitting MS were included in this open-label phase II trial. Subjects were observed without treatment for 6 months, followed by testosterone treatment for another 12 months. Focal gray matter loss as a marker for neurodegeneration was assessed using voxel-based morphometry. During the non-treatment phase, significant voxel-wise gray matter decreases were widespread (p≤ 0.05 corrected). However, during testosterone treatment, gray matter loss was no longer evident. In fact, a significant gray matter increase in the right frontal cortex was observed (p≤ 0.05 corrected). These observations support the potential of testosterone treatment to stall (and perhaps even reverse) neurodegeneration associated with MS. Furthermore, they warrant the investigation of testosterone's neuroprotective effects in larger, placebo controlled MS trials as well as in other neurodegenerative diseases. This is the first report of gray matter increase as the result of treatment in MS.

YNIMG Journal 2014 Journal Article

Why size matters: Differences in brain volume account for apparent sex differences in callosal anatomy

• Eileen Luders
• Arthur W. Toga
• Paul M. Thompson

Numerous studies have demonstrated a sexual dimorphism of the human corpus callosum. However, the question remains if sex differences in brain size, which typically is larger in men than in women, or biological sex per se account for the apparent sex differences in callosal morphology. Comparing callosal dimensions between men and women matched for overall brain size may clarify the true contribution of biological sex, as any observed group difference should indicate pure sex effects. We thus examined callosal morphology in 24 male and 24 female brains carefully matched for overall size. In addition, we selected 24 extremely large male brains and 24 extremely small female brains to explore if observed sex effects might vary depending on the degree to which male and female groups differed in brain size. Using the individual T1-weighted brain images (n=96), we delineated the corpus callosum at midline and applied a well-validated surface-based mesh-modeling approach to compare callosal thickness at 100 equidistant points between groups determined by brain size and sex. The corpus callosum was always thicker in men than in women. However, this callosal sex difference was strongly determined by the cerebral sex difference overall. That is, the larger the discrepancy in brain size between men and women, the more pronounced the sex difference in callosal thickness, with hardly any callosal differences remaining between brain-size matched men and women. Altogether, these findings suggest that individual differences in brain size account for apparent sex differences in the anatomy of the corpus callosum.

YNIMG Journal 2013 Journal Article

Decoding gender dimorphism of the human brain using multimodal anatomical and diffusion MRI data

• Delia-Lisa Feis
• Kay H. Brodersen
• D. Yves von Cramon
• Eileen Luders
• Marc Tittgemeyer

The female brain contains a larger proportion of gray matter tissue, while the male brain comprises more white matter. Findings like these have sparked increasing interest in studying dimorphism of the human brain: the general effect of gender on aspects of brain architecture. To date, the vast majority of imaging studies is based on unimodal MR images and typically limited to a small set of either gray or white matter regions-of-interest. The morphological content of magnetic resonance (MR) images, however, strongly depends on the underlying contrast mechanism. Consequently, in order to fully capture gender-specific morphological differences in distinct brain tissues, it might prove crucial to consider multiple imaging modalities simultaneously. This study introduces a novel approach to perform such multimodal classification incorporating the relative strengths of each modality-specific physical aperture to tissue properties. To illustrate our approach, we analyzed multimodal MR images (T1-, T2-, and diffusion-weighted) from 121 subjects (67 females) using a linear support vector machine with a mass-univariate feature selection procedure. We demonstrate that the combination of different imaging modalities yields a significantly higher balanced classification accuracy (96%) than any one modality by itself (83%–88%). Our results do not only confirm previous morphometric findings; crucially, they also shed new light on the most discriminative features in gray-matter volume and microstructure in cortical and subcortical areas. Specifically, we find that gender disparities are primarily distributed along brain networks thought to be involved in social cognition, reward-based learning, decision-making, and visual-spatial skills.

YNIMG Journal 2012 Journal Article

Brain maturation: Predicting individual BrainAGE in children and adolescents using structural MRI

• Katja Franke
• Eileen Luders
• Arne May
• Marko Wilke
• Christian Gaser

Background Neural development during human childhood and adolescence involves highly coordinated and sequenced events, characterized by both progressive and regressive processes. Despite a multitude of results demonstrating the age-dependent development of gray matter, white matter, and total brain volume, a reference curve allowing prediction of structural brain maturation is still lacking but would be clinically valuable. For the first time, the present study provides a validated reference curve for structural brain maturation during childhood and adolescence, based on structural MRI data. Methods and findings By employing kernel regression methods, a novel but well-validated BrainAGE framework uses the complex multidimensional maturation pattern across the whole brain to estimate an individual's brain age. The BrainAGE framework was applied to a large human sample (n=394) of healthy children and adolescents, whose image data had been acquired during the NIH MRI study of normal brain development. Using this approach, we were able to predict individual brain maturation with a clinically meaningful accuracy: the correlation between predicted brain age and chronological age resulted in r=0. 93. The mean absolute error was only 1. 1years. Moreover, the predicted brain age reliably differentiated between all age groups (i. e. , preschool childhood, late childhood, early adolescence, middle adolescence, late adolescence). Applying the framework to preterm-born adolescents resulted in a significantly lower estimated brain age than chronological age in subjects who were born before the end of the 27th week of gestation, demonstrating the successful clinical application and future potential of this method. Conclusions Consequently, in the future this novel BrainAGE approach may prove clinically valuable in detecting both normal and abnormal brain maturation, providing important prognostic information.

YNIMG Journal 2012 Journal Article

Bridging the hemispheres in meditation: Thicker callosal regions and enhanced fractional anisotropy (FA) in long-term practitioners

• Eileen Luders
• Owen R. Phillips
• Kristi Clark
• Florian Kurth
• Arthur W. Toga
• Katherine L. Narr

Recent findings suggest a close link between long-term meditation practices and the structure of the corpus callosum. Prior analyses, however, have focused on estimating mean fractional anisotropy (FA) within two large pre-defined callosal tracts only. Additional effects might exist in other, non-explored callosal regions and/or with respect to callosal attributes not captured by estimates of FA. To further explore callosal features in the framework of meditation, we analyzed 30 meditators and 30 controls, carefully matched for sex, age, and handedness. We applied a multimodal imaging approach using diffusion tensor imaging (DTI) in combination with structural magnetic resonance imaging (MRI). Callosal measures of tract-specific FA were complemented with other global (segment-specific) estimates as well as extremely local (point-wise) measures of callosal micro- and macro-structure. Callosal measures were larger in long-term meditators compared to controls, particularly in anterior callosal sections. However, differences achieved significance only when increasing the regional sensitivity of the measurement (i. e. , using point-wise measures versus segment-specific measures) and were more prominent for microscopic than macroscopic characteristics (i. e. , callosal FA versus callosal thickness). Thicker callosal regions and enhanced FA in meditators might indicate greater connectivity, possibly reflecting increased hemispheric integration during cerebral processes involving (pre)frontal regions. Such a brain organization might be linked to achieving characteristic mental states and skills as associated with meditation, though this hypothesis requires behavioral confirmation. Moreover, longitudinal studies are required to address whether the observed callosal effects are induced by meditation or constitute an innate prerequisite for the start or successful continuation of meditation.

YNIMG Journal 2011 Journal Article

Enhanced brain connectivity in long-term meditation practitioners

• Eileen Luders
• Kristi Clark
• Katherine L. Narr
• Arthur W. Toga

Very little is currently known about the cerebral characteristics that underlie the complex processes of meditation as only a limited number of studies have addressed this topic. Research exploring structural connectivity in meditation practitioners is particularly rare. We thus acquired diffusion tensor imaging (DTI) data of high angular and spatial resolution and used atlas-based tract mapping methods to investigate white matter fiber characteristics in a well-matched sample of long-term meditators and controls (n=54). A broad field mapping approach estimated the fractional anisotropy (FA) for twenty different fiber tracts (i. e. , nine tracts in each hemisphere and two inter-hemispheric tracts) that were subsequently used as dependent measures. Results showed pronounced structural connectivity in meditators compared to controls throughout the entire brain within major projection pathways, commissural pathways, and association pathways. The largest group differences were observed within the corticospinal tract, the temporal component of the superior longitudinal fasciculus, and the uncinate fasciculus. While cross-sectional studies represent a good starting point for elucidating possible links between meditation and white matter fiber characteristics, longitudinal studies will be necessary to determine the relative contribution of nature and nurture to enhanced structural connectivity in long-term meditators.

YNIMG Journal 2011 Journal Article

The link between callosal thickness and intelligence in healthy children and adolescents

• Eileen Luders
• Paul M. Thompson
• Katherine L. Narr
• Alen Zamanyan
• Yi-Yu Chou
• Boris Gutman
• Ivo D. Dinov
• Arthur W. Toga

The link between brain structure and intelligence is a well-investigated topic, but existing analyses have mainly focused on adult samples. Studies in healthy children and adolescents are rare, and normative data specifically addressing the association between corpus callosum morphology and intellectual abilities are quite limited. To advance this field of research, we mapped the correlations between standardized intelligence measures and callosal thickness based on high-resolution magnetic resonance imaging (MRI) data. Our large and well-matched sample included 200 normally developing subjects (100 males, 100 females) ranging from 6 to 17years of age. Although the strongest correlations were negative and confined to the splenium, the strength and the direction of intelligence-callosal thickness associations varied considerably. While significant correlations in females were mainly positive, significant correlations in males were exclusively negative. However, only the negative correlations in the overall sample (i. e. , males and females combined) remained significant when controlling for multiple comparisons. The observed negative correlations between callosal thickness and intelligence in children and adolescents contrast with the positive correlations typically reported in adult samples. However, negative correlations are in line with reports from other pediatric studies relating cognitive measures to other brain attributes such as cortical thickness, gray matter volume, and gray matter density. Altogether, these findings suggest that relationships between callosal morphology and cognition are highly dynamic during brain maturation. Sex effects on links between callosal thickness and intelligence during childhood and adolescence are present but appear rather weak in general.

YNIMG Journal 2010 Journal Article

Callosal atrophy in mild cognitive impairment and Alzheimer's disease: Different effects in different stages

• Margherita Di Paola
• Eileen Luders
• Fulvia Di Iulio
• Andrea Cherubini
• Domenico Passafiume
• Paul M. Thompson
• Carlo Caltagirone
• Arthur W. Toga

Alzheimer's Disease (AD) is a neurodegenerative disorder that mainly affects grey matter (GM). Nevertheless, a number of investigations have documented white matter (WM) pathology associated with AD. The corpus callosum (CC) is the largest WM fiber bundle in the human brain. It has been shown to be susceptible to atrophy in AD mainly as a correlate of Wallerian degeneration of commissural nerve fibers of the neocortex. The aim of this study was to investigate which callosal regions are affected and whether callosal degeneration is associated with the stage of the disease. For this purpose, we analyzed high-resolution MRI data of patients with amnesic mild cognitive impairment (MCI) (n =20), mild AD (n =20), severe AD (n =10), and of healthy controls (n =20). Callosal morphology was investigated applying two different structural techniques: mesh-based geometrical modeling methods and whole-brain voxel-based analyses. Our findings indicate significant reductions in severe AD patients compared to healthy controls in anterior (genu and anterior body) and posterior (splenium) sections. In contrast, differences between healthy controls and mild AD patients or amnesic MCI patients were less pronounced and did not survive corrections for multiple comparisons. When correlating anterior and posterior WM density of the CC with GM density of the cortex in the severe AD group, we detected significant positive relationships between posterior sections of the CC and the cortex. We conclude that callosal atrophy is present predominantly in the latest stage of AD, where two mechanisms might contribute to WM alterations in severe AD: the Wallerian degeneration in posterior subregions and the myelin breakdown process in anterior subregions.

YNIMG Journal 2010 Journal Article

When more is less: Associations between corpus callosum size and handedness lateralization

• Eileen Luders
• Nicolas Cherbuin
• Paul M. Thompson
• Boris Gutman
• Kaarin J. Anstey
• Perminder Sachdev
• Arthur W. Toga

Although not consistently replicated, a substantial number of studies suggest that left-handers have larger callosal regions than right-handers. We challenge this notion and propose that callosal size is not linked to left-handedness or right-handedness per se but to the degree of handedness lateralization. To test this hypothesis, we investigated the thickness of the corpus callosum in a large data set (n =361). We analyzed the correlations between callosal thickness and the degree of handedness lateralization in 324 right-handers and 37 left-handers at 100 equidistant points across the corpus callosum. We revealed significant negative correlations within the anterior and posterior midbody suggesting that larger callosal dimensions in these regions are associated with a weaker handedness lateralization. Significant positive correlations were completely absent. In addition, we compared callosal thickness between moderately lateralized left-handers (n =37) and three equally sized groups (n =37) of right-handers (strongly, moderately, and weakly lateralized). The outcomes of these group analyses confirmed the negative association between callosal size and handedness lateralization, although callosal differences between right- and left-handers did not reach statistical significance. This suggests that callosal differences are rather small, if examined as a dichotomy between two handedness groups. Future studies will expand this line of research by increasing the number of left-handers to boost statistical power and by combining macro- and microstructural, as well as functional and behavioral measurements to identify the biological mechanisms linking callosal morphology and handedness lateralization.

YNIMG Journal 2009 Journal Article

Regional gray matter variation in male-to-female transsexualism

• Eileen Luders
• Francisco J. Sánchez
• Christian Gaser
• Arthur W. Toga
• Katherine L. Narr
• Liberty S. Hamilton
• Eric Vilain

Gender identity—one's sense of being a man or a woman—is a fundamental perception experienced by all individuals that extends beyond biological sex. Yet, what contributes to our sense of gender remains uncertain. Since individuals who identify as transsexual report strong feelings of being the opposite sex and a belief that their sexual characteristics do not reflect their true gender, they constitute an invaluable model to understand the biological underpinnings of gender identity. We analyzed MRI data of 24 male-to-female (MTF) transsexuals not yet treated with cross-sex hormones in order to determine whether gray matter volumes in MTF transsexuals more closely resemble people who share their biological sex (30 control men), or people who share their gender identity (30 control women). Results revealed that regional gray matter variation in MTF transsexuals is more similar to the pattern found in men than in women. However, MTF transsexuals show a significantly larger volume of regional gray matter in the right putamen compared to men. These findings provide new evidence that transsexualism is associated with distinct cerebral pattern, which supports the assumption that brain anatomy plays a role in gender identity.

YNIMG Journal 2009 Journal Article

The underlying anatomical correlates of long-term meditation: Larger hippocampal and frontal volumes of gray matter

• Eileen Luders
• Arthur W. Toga
• Natasha Lepore
• Christian Gaser

Although the systematic study of meditation is still in its infancy, research has provided evidence for meditation-induced improvements in psychological and physiological well-being. Moreover, meditation practice has been shown not only to benefit higher-order cognitive functions but also to alter brain activity. Nevertheless, little is known about possible links to brain structure. Using high-resolution MRI data of 44 subjects, we set out to examine the underlying anatomical correlates of long-term meditation with different regional specificity (i. e. , global, regional, and local). For this purpose, we applied voxel-based morphometry in association with a recently validated automated parcellation approach. We detected significantly larger gray matter volumes in meditators in the right orbito-frontal cortex (as well as in the right thalamus and left inferior temporal gyrus when co-varying for age and/or lowering applied statistical thresholds). In addition, meditators showed significantly larger volumes of the right hippocampus. Both orbito-frontal and hippocampal regions have been implicated in emotional regulation and response control. Thus, larger volumes in these regions might account for meditators' singular abilities and habits to cultivate positive emotions, retain emotional stability, and engage in mindful behavior. We further suggest that these regional alterations in brain structures constitute part of the underlying neurological correlate of long-term meditation independent of a specific style and practice. Future longitudinal analyses are necessary to establish the presence and direction of a causal link between meditation practice and brain anatomy.

YNIMG Journal 2008 Journal Article

Erratum to “A meta-algorithm for brain extraction in MRI” [NeuroImage 23 (2004) 625–637]

• David E. Rex
• David W. Shattuck
• Roger P. Woods
• Katherine L. Narr
• Eileen Luders
• Kelly Rehm
• Sarah E. Stoltzner
• David A. Rottenberg

YNIMG Journal 2007 Journal Article

Asymmetries of cortical shape: Effects of handedness, sex and schizophrenia

• Katherine L. Narr
• Robert M. Bilder
• Eileen Luders
• Paul M. Thompson
• Roger P. Woods
• Delbert Robinson
• Philip R. Szeszko
• Teodora Dimtcheva

Some evidence suggests that sex, handedness and disease processes associated with schizophrenia affect the magnitude and/or direction of structural brain asymmetries. There are mixed findings, however, on how these factors influence cerebral torque, when torque is assessed with linear or volumetric measurements. We obtained MRI data from 67 healthy (30 males, 10 non-dextrals) and 84 schizophrenia subjects (60 males; 16 non-dextrals) and applied cortical pattern matching to spatially relate and compare differences in the surface morphology of the two cerebral hemispheres at high spatial resolution. Asymmetry indices, computed at thousands of matched hemispheric locations, were used to examine effects of sex, handedness and schizophrenia on hemispheric shape asymmetries while controlling for age and the other factors. Highly significant and discriminative right-frontal and left parietal–occipital surface expansions and protrusions (petalias) were mapped within groups. Although hemispheric shape asymmetries appeared less pronounced within female non-dextrals, asymmetry indices were not shown to differ significantly across sex, hand preference or diagnosis, or to reveal interactions of handedness with sex or diagnosis. Our 3D maps of spatially detailed anterior and posterior hemispheric shape asymmetries reflect subtle geometric distortions in hemispheric surface morphology that cannot be characterized with 2D or volumetric methods. Inter-individual variations in hemispheric torque appear minimally influenced by sex, dextrality or disease status. Biological factors driving language dominance or other lateralized brain functions dissociable from handedness, may more closely relate to hemispheric shape asymmetries, while the lateralization of other discrete brain regions may be more influenced by sexually dimorphic factors or by schizophrenia pathophysiology.

YNIMG Journal 2007 Journal Article

Positive correlations between corpus callosum thickness and intelligence

• Eileen Luders
• Katherine L. Narr
• Robert M. Bilder
• Paul M. Thompson
• Philip R. Szeszko
• Liberty Hamilton
• Arthur W. Toga

Callosal morphology is thought to reflect the capacity for inter-hemispheric communication and thus, in addition to other cerebral characteristics, may serve as a neuroanatomical substrate of general intellectual capacity. We applied novel computational mesh-based methods to establish the presence and direction of correlations between intelligence and callosal thickness at high spatial resolution while removing the variance associated with overall brain size. Within healthy subjects (n =62), and within males (n =28) and females (n =34) separately, we observed significant positive correlations between callosal morphology and intelligence measures (full-scale, performance, and verbal). These relationships were pronounced in posterior callosal sections and were confirmed by permutation testing. Significant negative correlations were absent. Positive associations between intelligence and posterior callosal thickness may reflect a more efficient inter-hemispheric information transfer, positively affecting information processing and integration, and thus intellectual performance. At the same time, regional variations in callosal size might also partly reflect the underlying architecture of topographically connected cortical regions relevant for processing higher-order cognitive information. Our findings emphasize the importance of incorporating posterior (callosal) regions into the theories and models proposed to explain the anatomical substrates of intelligence.

YNIMG Journal 2006 Journal Article

Increased local gyrification mapped in Williams syndrome

• Christian Gaser
• Eileen Luders
• Paul M. Thompson
• Agatha D. Lee
• Rebecca A. Dutton
• Jennifer A. Geaga
• Kiralee M. Hayashi
• Ursula Bellugi

Applying a recently developed method to analyze gyrification with excellent spatial resolution across thousands of points across the lateral and medial cortical surface, we mapped differences in cortical surface anatomy between subjects with Williams syndrome (WS; n =42) and an age-matched sample of healthy subjects (n =40). WS subjects showed increased gyrification bilaterally in occipital regions and over the cuneus. Differences were more pronounced in the left hemisphere than in the right, with additional regions of increased gyrification in WS in the left precuneus, posterior and anterior cingulate, paracentral and mesial frontal lobe. No cortical area was significantly more convoluted in healthy subjects relative to the WS subjects. On the lateral surfaces, the direction and pattern of gyrification asymmetries were similar in WS subjects and controls; posterior brain regions had greater gyrification in the left hemisphere, while anterior brain regions showed greater gyrification in the right hemisphere. On the medial surfaces, control subjects and WS individuals differed considerably with respect to the degree but also direction of gyrification asymmetry. Our findings confirm and extend previous studies measuring cortical complexity at the global whole-brain or hemispheric levels. The observed gyrification abnormalities in individuals with WS might be related to dysfunctions in neuronal circuits and consequently contribute to the distinct cognitive and behavioral profile accompanying the disorder.

YNIMG Journal 2004 Journal Article

A meta-algorithm for brain extraction in MRI

• David E. Rex
• David W. Shattuck
• Roger P. Woods
• Katherine L. Narr
• Eileen Luders
• Kelly Rehm
• Sarah E. Stolzner
• David A. Rottenberg

Accurate identification of brain tissue and cerebrospinal fluid (CSF) in a whole-head MRI is a critical first step in many neuroimaging studies. Automating this procedure can eliminate intra- and interrater variance and greatly increase throughput for a labor-intensive step. Many available procedures perform differently across anatomy and under different acquisition protocols. We developed the Brain Extraction Meta-Algorithm (BEMA) to address these concerns. It executes many extraction algorithms and a registration procedure in parallel to combine the results in an intelligent fashion and obtain improved results over any of the individual algorithms. Using an atlas space, BEMA performs a voxelwise analysis of training data to determine the optimal Boolean combination of extraction algorithms to produce the most accurate result for a given voxel. This allows the provided extractors to be used differentially across anatomy, increasing both the accuracy and robustness of the procedure. We tested BEMA using modified forms of BrainSuite's Brain Surface Extractor (BSE), FSL's Brain Extraction Tool (BET), AFNI's 3dIntracranial, and FreeSurfer's MRI Watershed as well as FSL's FLIRT for the registration procedure. Training was performed on T1-weighted scans of 136 subjects from five separate data sets with different acquisition parameters on separate scanners. Testing was performed on 135 separate subjects from the same data sets. BEMA outperformed the individual algorithms, as well as interrater results from a subset of the scans, when compared for the mean Dice coefficient, a rating of the similarity of output masks to the manually defined gold standards.