Min Tae M. Park

Multi-atlas segmentation of the whole hippocampus and subfields using multiple automatically generated templates

INTRODUCTION Advances in image segmentation of magnetic resonance images (MRI) have demonstrated that multi-atlas approaches improve segmentation over regular atlas-based approaches. These approaches often rely on a large number of manually segmented atlases (e.g. 30-80) that take significant time and expertise to produce. We present an algorithm, MAGeT-Brain (Multiple Automatically Generated Templates), for the automatic segmentation of the hippocampus that minimises the number of atlases needed whilst still achieving similar agreement to multi-atlas approaches. Thus, our method acts as a reliable multi-atlas approach when using special or hard-to-define atlases that are laborious to construct. METHOD MAGeT-Brain works by propagating atlas segmentations to a template library, formed from a subset of target images, via transformations estimated by nonlinear image registration. The resulting segmentations are then propagated to each target image and fused using a label fusion method. We conduct two separate Monte Carlo cross-validation experiments comparing MAGeT-Brain and basic multi-atlas whole hippocampal segmentation using differing atlas and template library sizes, and registration and label fusion methods. The first experiment is a 10-fold validation (per parameter setting) over 60 subjects taken from the Alzheimers Disease Neuroimaging Database (ADNI), and the second is a five-fold validation over 81 subjects having had a first episode of psychosis. In both cases, automated segmentations are compared with manual segmentations following the Pruessner-protocol. Using the best settings found from these experiments, we segment 246 images of the ADNI1:Complete 1Yr 1.5 T dataset and compare these with segmentations from existing automated and semi-automated methods: FSL FIRST, FreeSurfer, MAPER, and SNT. Finally, we conduct a leave-one-out cross-validation of hippocampal subfield segmentation in standard 3T T1-weighted images, using five high-resolution manually segmented atlases (Winterburn et al., 2013). RESULTS In the ADNI cross-validation, using 9 atlases MAGeT-Brain achieves a mean Dices Similarity Coefficient (DSC) score of 0.869 with respect to manual whole hippocampus segmentations, and also exhibits significantly lower variability in DSC scores than multi-atlas segmentation. In the younger, psychosis dataset, MAGeT-Brain achieves a mean DSC score of 0.892 and produces volumes which agree with manual segmentation volumes better than those produced by the FreeSurfer and FSL FIRST methods (mean difference in volume: 80 mm(3), 1600 mm(3), and 800 mm(3), respectively). Similarly, in the ADNI1:Complete 1Yr 1.5 T dataset, MAGeT-Brain produces hippocampal segmentations well correlated (r>0.85) with SNT semi-automated reference volumes within disease categories, and shows a conservative bias and a mean difference in volume of 250 mm(3) across the entire dataset, compared with FreeSurfer and FSL FIRST which both overestimate volume differences by 2600 mm(3) and 2800 mm(3) on average, respectively. Finally, MAGeT-Brain segments the CA1, CA4/DG and subiculum subfields on standard 3T T1-weighted resolution images with DSC overlap scores of 0.56, 0.65, and 0.58, respectively, relative to manual segmentations. CONCLUSION We demonstrate that MAGeT-Brain produces consistent whole hippocampal segmentations using only 9 atlases, or fewer, with various hippocampal definitions, disease populations, and image acquisition types. Additionally, we show that MAGeT-Brain identifies hippocampal subfields in standard 3T T1-weighted images with overlap scores comparable to competing methods.

Illness Progression, Recent Stress, and Morphometry of Hippocampal Subfields and Medial Prefrontal Cortex in Major Depression

BACKGROUND Longitudinal studies of illness progression in patients with major depressive disorder (MDD) indicate that the onset of subsequent depressive episodes becomes increasingly decoupled from external stressors. A possible mechanism underlying this phenomenon is that multiple episodes induce long-lasting neurobiological changes that confer increased risk for recurrence. Prior morphometric studies have frequently reported volumetric reductions in patients with MDD--especially in medial prefrontal cortex (mPFC) and the hippocampus--but few studies have investigated whether these changes are exacerbated by prior episodes. METHODS In a sample of 103 medication-free patients with depression and control subjects with no history of depression, structural magnetic resonance imaging was performed to examine relationships between number of prior episodes, current stress, hippocampal subfield volume and cortical thickness. Volumetric analyses of the hippocampus were performed using a recently validated subfield segmentation approach, and cortical thickness estimates were obtained using vertex-based methods. Participants were grouped on the basis of the number of prior depressive episodes and current depressive diagnosis. RESULTS Number of prior episodes was associated with both lower reported stress levels and reduced volume in the dentate gyrus. Cortical thinning of the left mPFC was associated with a greater number of prior depressive episodes but not current depressive diagnosis. CONCLUSIONS Collectively, these findings are consistent with preclinical models suggesting that the dentate gyrus and mPFC are especially vulnerable to stress exposure and provide evidence for morphometric changes that are consistent with stress-sensitization models of recurrence in MDD.

Neonatal Pain and Infection Relate to Smaller Cerebellum in Very Preterm Children at School Age

OBJECTIVE To examine whether specific neonatal factors differentially influence cerebellar subregional volumes and to investigate relationships between subregional volumes and outcomes in very preterm children at 7 years of age. STUDY DESIGN Fifty-six children born very preterm (24-32 weeks gestational age) followed longitudinally from birth underwent 3-dimensional T(1)-weighted neuroimaging at median age 7.6 years. Children with severe brain injury were excluded. Cerebellar subregions were automatically segmented using the multiple automatically generated templates algorithm. The relation between cerebellum subregional volumes (adjusted for total brain volume and sex) and neonatal clinical factors were examined using constrained principal component analysis. Cognitive and visual-motor integration functions in relation to cerebellar volumes were also investigated. RESULTS Higher neonatal procedural pain and infection, as well as other clinical factors, were differentially associated with reduced cerebellar volumes in specific subregions. After adjusting for clinical risk factors, neonatal procedural pain was distinctively associated with smaller volumes bilaterally in the posterior VIIIA and VIIIB lobules. Specific smaller cerebellar subregional volumes were related to poorer cognition and motor/visual integration. CONCLUSIONS In very preterm children, exposure to painful procedures, as well as additional neonatal risk factors such as infection, were associated with reduced cerebellar volumes in specific subregions and poorer outcomes at school age.

Morphological Alterations in the Thalamus, Striatum, and Pallidum in Autism Spectrum Disorder.

Autism spectrum disorder (ASD) is a common neurodevelopmental disorder with cognitive, motor, and emotional symptoms. The thalamus and basal ganglia form circuits with the cortex supporting all three of these behavioral domains. Abnormalities in the structure of subcortical regions may suggest atypical development of these networks, with implications for understanding the neural basis of ASD symptoms. Findings from previous volumetric studies have been inconsistent. Here, using advanced surface-based methodology, we investigated localized differences in shape and surface area in the basal ganglia and thalamus in ASD, using T1-weighted anatomical images from the Autism Brain Imaging Data Exchange (373 male participants aged 7–35 years with ASD and 384 typically developing). We modeled effects of diagnosis, age, and their interaction on volume, shape, and surface area. In participants with ASD, we found expanded surface area in the right posterior thalamus corresponding to the pulvinar nucleus, and a more concave shape in the left mediodorsal nucleus. The shape of both caudal putamen and pallidum showed a relatively steeper increase in concavity with age in ASD. Within ASD participants, restricted, repetitive behaviors were positively associated with surface area in bilateral globus pallidus. We found no differences in overall volume, suggesting that surface-based approaches have greater sensitivity to detect localized differences in subcortical structure. This work adds to a growing body of literature implicating corticobasal ganglia-thalamic circuits in the pathophysiology of ASD. These circuits subserve a range of cognitive, emotional, and motor functions, and may have a broad role in the complex symptom profile in ASD.

Striatal Morphology is Associated with Tobacco Cigarette Craving

The striatum has a clear role in addictive disorders and is involved in drug-related craving. Recently, enhanced striatal volume was associated with greater lifetime nicotine exposure, suggesting a bridge between striatal function and structural phenotypes. To assess this link between striatal structure and function, we evaluated the relationship between striatal morphology and this brain region’s well-established role in craving. In tobacco smokers, we assessed striatal volume, surface area, and shape using a new segmentation methodology coupled with local shape indices. Striatal morphology was then related with two measures of craving: state-based craving, assessed by the brief questionnaire of smoking urges (QSU), and craving induced by smoking-related images. A positive association was found between left striatal volume and surface area with both measures of craving. A more specific relationship was found between both craving measures and the dorsal, but not in ventral striatum. Evaluating dorsal striatal subregions showed a single relationship between the caudate and QSU. Although cue-induced craving and the QSU were both associated with enlarged striatal volume and surface area, these measures were differentially associated with global or more local striatal volumes. We also report a connection between greater right striatal shape deformations and cue-induced craving. Shape deformations associated with cue-induced craving were specific to striatal subregions involved in habitual responding to rewarding stimuli, which is relevant given the habitual nature of cue-induced craving. The current findings confirm a relationship between striatal function and morphology and suggest that variation in striatal morphology may be a biomarker for craving severity.

Regional cerebellar volumes are related to early musical training and finger tapping performance

The cerebellum has been associated with timing on the millisecond scale and with musical rhythm and beat processing. Early musical training (before age 7) is associated with enhanced rhythm synchronization performance and differences in cortical motor areas and the corpus callosum. In the present study, we examined the relationships between regional cerebellar volumes, early musical training, and timing performance. We tested adult musicians and non-musicians on a standard finger tapping task, and extracted cerebellar gray and white matter volumes using a novel multi-atlas automatic segmentation pipeline. We found that early-trained musicians had reduced volume in bilateral cerebellar white matter and right lobules IV, V and VI, compared to late-trained musicians. Strikingly, better timing performance, greater musical experience and an earlier age of start of musical training were associated with smaller cerebellar volumes. Better timing performance was specifically associated with smaller volumes of right lobule VI. Collectively, these findings support the sensitivity of the cerebellum to the age of initiation of musical training and suggest that lobule VI plays a role in timing. The smaller cerebellar volumes associated with musical training and timing performance may be a reflection of more efficiently implemented low-level timing and sensorimotor processes.

Estimating volumes of the pituitary gland from T1-weighted magnetic-resonance images: effects of age, puberty, testosterone, and estradiol.

The pituitary gland is a key structure in the hypothalamic-pituitary-gonadal (HPG) axis--it plays an important role in sexual maturation during puberty. Despite its small size, its volume can be quantified using magnetic resonance imaging (MRI). Here, we study a cohort of 962 typically developing adolescents from the Saguenay Youth Study and estimate pituitary volumes using a newly developed multi-atlas segmentation method known as the MAGeT Brain algorithm. We found that age and puberty stage (controlled for age) each predicts adjusted pituitary volumes (controlled for total brain volume) in both males and females. Controlling for the effects of age and puberty stage, total testosterone and estradiol levels also predict adjusted pituitary volumes in males and pre-menarche females, respectively. These findings demonstrate that the pituitary gland grows during adolescence, and its volume relates to circulating plasma-levels of sex steroids in both males and females.

CERES: A new cerebellum lobule segmentation method

ABSTRACT The human cerebellum is involved in language, motor tasks and cognitive processes such as attention or emotional processing. Therefore, an automatic and accurate segmentation method is highly desirable to measure and understand the cerebellum role in normal and pathological brain development. In this work, we propose a patch‐based multi‐atlas segmentation tool called CERES (CEREbellum Segmentation) that is able to automatically parcellate the cerebellum lobules. The proposed method works with standard resolution magnetic resonance T1‐weighted images and uses the Optimized PatchMatch algorithm to speed up the patch matching process. The proposed method was compared with related recent state‐of‐the‐art methods showing competitive results in both accuracy (average DICE of 0.7729) and execution time (around 5 minutes). HIGHLIGHTSWe present a novel method for cerebellum lobule segmentation on MRI.The method consists of a fast multi‐atlas non‐local patch‐based label fusion.Our proposed method was shown to improve the state‐of‐the‐art methods with a reduced temporal cost (5 minutes).The pipeline presented in this work will be made available to scientific community through our web ‐based platform volBrain.

Functional and Structural Correlates of Memory in Patients with Mesial Temporal Lobe Epilepsy

Individuals with medial temporal lobe epilepsy (mTLE) often show material-specific memory impairment (verbal for left, visuospatial for right hemisphere), which can be exacerbated following surgery aimed at the epileptogenic regions of medial and anterolateral temporal cortex. There is a growing body of evidence suggesting that characterization of structural and functional integrity of these regions using MRI can aid in prediction of post-surgical risk of further memory decline. We investigated the nature of the relationship between structural and functional indices of hippocampal integrity with pre-operative memory performance in a group of 26 patients with unilateral mTLE. Structural integrity was assessed using hippocampal volumes, while functional integrity was assessed using hippocampal activation during the encoding of novel scenes. We quantified structural and functional integrity in terms of asymmetry, calculated as (L − R)/(L + R). Factor scores for verbal and visual memory were calculated from a clinical database and an asymmetry score (verbal − visual) was used to characterize memory performance. We found, as expected, a significant difference between left and right mTLE (RTLE) groups for hippocampal volume asymmetry, with each group showing an asymmetry favoring the unaffected temporal lobe. Encoding activation asymmetry showed a similar pattern, with left mTLE patients showing activation preferential to the right hemisphere and RTLE patients showing the reverse. Finally, we demonstrated that functional integrity mediated the relationship between structural integrity and memory performance for memory asymmetry, suggesting that even if structural changes are evident, ultimately it is the functional integrity of the tissue that most closely explains behavioral performance.

Manual segmentation of the fornix, fimbria, and alveus on high-resolution 3T MRI: Application via fully-automated mapping of the human memory circuit white and grey matter in healthy and pathological aging.

ABSTRACT Recently, much attention has been focused on the definition and structure of the hippocampus and its subfields, while the projections from the hippocampus have been relatively understudied. Here, we derive a reliable protocol for manual segmentation of hippocampal white matter regions (alveus, fimbria, and fornix) using high‐resolution magnetic resonance images that are complementary to our previous definitions of the hippocampal subfields, both of which are freely available at https://github.com/cobralab/atlases. Our segmentation methods demonstrated high inter‐ and intra‐rater reliability, were validated as inputs in automated segmentation, and were used to analyze the trajectory of these regions in both healthy aging (OASIS), and Alzheimers disease (AD) and mild cognitive impairment (MCI; using ADNI). We observed significant bilateral decreases in the fornix in healthy aging while the alveus and cornu ammonis (CA) 1 were well preserved (all ps<0.006). MCI and AD demonstrated significant decreases in fimbriae and fornices. Many hippocampal subfields exhibited decreased volume in both MCI and AD, yet no significant differences were found between MCI and AD cohorts themselves. Our results suggest a neuroprotective or compensatory role for the alveus and CA1 in healthy aging and suggest that an improved understanding of the volumetric trajectories of these structures is required. HIGHLIGHTSNovel high‐resolution manual segmentation of human alveus, fimbria, and fornix.Validation (precision and accuracy) of manual atlases for use in automatic segmentation.Application of automatic segmentation on AD/MCI and healthy aging datasets.Results suggest neuroprotective role for alveus and hippocampal CA1 region.