Comparison of tissue segmentation algorithms in neuroimage analysis software tools

Insights from auditory cortex for GABA+ magnetic resonance spectroscopy studies of aging

Changes in levels of the inhibitory neurotransmitter γ-aminobutyric acid (GABA) may underlie aging-related changes in brain function. GABA and co-edited macromolecules (GABA+) can be measured with MEGA-PRESS magnetic resonance spectroscopy (MRS). The current study investigated how changes in the aging brain impact the interpretation of GABA+ measures in bilateral auditory cortices of healthy young and older adults. Structural changes during aging appeared as decreasing proportion of grey matter in the MRS volume of interest and corresponding increase in cerebrospinal fluid. GABA+ referenced to H₂ O without tissue correction declined in aging. This decline persisted after correcting for tissue differences in MR-visible H₂ O and relaxation times but vanished after considering the different abundance of GABA+ in grey and white matter. However, GABA+ referenced to creatine and N-acetyl aspartate (NAA), which showed no dependence on tissue composition, decreased in aging. All GABA+ measures showed hemispheric asymmetry in young but not older adults. The study also considered aging-related effects on tissue segmentation and the impact of co-edited macromolecules. Tissue segmentation differed significantly between commonly used algorithms, but aging-related effects on tissue-corrected GABA+ were consistent across methods. Auditory cortex macromolecule concentration did not change with age, indicating that a decline in GABA caused the decrease in the compound GABA+ measure. Most likely, the macromolecule contribution to GABA+ leads to underestimating an aging-related decrease in GABA. Overall, considering multiple GABA+ measures using different reference signals strengthened the support for an aging-related decline in auditory cortex GABA levels.

Validation of an automatic tool for the rapid measurement of brain atrophy and white matter hyperintensity: QyScore®

OBJECTIVES

QyScore® is an imaging analysis tool certified in Europe (CE marked) and the US (FDA cleared) for the automatic volumetry of grey and white matter (GM and WM respectively), hippocampus (HP), amygdala (AM), and white matter hyperintensity (WMH). Here we compare QyScore® performances with the consensus of expert neuroradiologists.

METHODS

Dice similarity coefficient (DSC) and the relative volume difference (RVD) for GM, WM volumes were calculated on 50 3DT1 images. DSC and the F1 metrics were calculated for WMH on 130 3DT1 and FLAIR images. For each index, we identified thresholds of reliability based on current literature review results. We hypothesized that DSC/F1 scores obtained using QyScore® markers would be higher than the threshold. In contrast, RVD scores would be lower. Regression analysis and Bland-Altman plots were obtained to evaluate QyScore® performance in comparison to the consensus of three expert neuroradiologists.

RESULTS

The lower bound of the DSC/F1 confidence intervals was higher than the threshold for the GM, WM, HP, AM, and WMH, and the higher bounds of the RVD confidence interval were below the threshold for the WM, GM, HP, and AM. QyScore®, compared with the consensus of three expert neuroradiologists, provides reliable performance for the automatic segmentation of the GM and WM volumes, and HP and AM volumes, as well as WMH volumes.

CONCLUSIONS

QyScore® represents a reliable medical device in comparison with the consensus of expert neuroradiologists. Therefore, QyScore® could be implemented in clinical trials and clinical routine to support the diagnosis and longitudinal monitoring of neurological diseases.

KEY POINTS

• QyScore® provides reliable automatic segmentation of brain structures in comparison with the consensus of three expert neuroradiologists. • QyScore® automatic segmentation could be performed on MRI images using different vendors and protocols of acquisition. In addition, the fast segmentation process saves time over manual and semi-automatic methods. • QyScore® could be implemented in clinical trials and clinical routine to support the diagnosis and longitudinal monitoring of neurological diseases.

Dispositional Negative Emotionality in Childhood and Adolescence Predicts Structural Variation in the Amygdala and Caudal Anterior Cingulate During Early Adulthood: Theoretically and Empirically Based Tests

Substantial evidence implicates the amygdala and related structures in the processing of negative emotions. Furthermore, neuroimaging evidence suggests that variations in amygdala volumes are related to trait-like individual differences in neuroticism/negative emotionality, although many questions remain about the nature of such associations. We conducted planned tests of the directional prediction that dispositional negative emotionality measured at 10-17 years using parent and youth ratings on the Child and Adolescent Dispositions Scale (CADS) would predict larger volumes of the amygdala in adulthood and conducted exploratory tests of associations with other regions implicated in emotion processing. Participants were 433 twins strategically selected for neuroimaging during wave 2 from wave 1 of the Tennessee Twins Study (TTS) by oversampling on internalizing and/or externalizing psychopathology risk. Controlling for age, sex, race-ethnicity, handedness, scanner, and total brain volume, youth-rated negative emotionality positively predicted bilateral amygdala volumes after correction for multiple testing. Each unit difference of one standard deviation (SD) in negative emotionality was associated with a .12 SD unit difference in larger volumes of both amygdalae. Parent-rated negative emotionality predicted greater thickness of the left caudal/dorsal anterior cingulate cortex (β = 0.28). Associations of brain structure with negative emotionality were not moderated by sex. These results are striking because dispositions assessed at 10-17 years of age were predictive of grey matter volumes measured 12-13 years later in adulthood. Future longitudinal studies should examine the timing of amygdala/cingulate associations with dispositional negative emotionality to determine when these associations emerge during development.

Fast and accurate pseudo multispectral technique for whole-brain MRI tissue classification

Numerous strategies have been proposed to classify brain tissues into gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF). However, many of them fail when classifying specific regions with low contrast between tissues. In this work, we propose an alternative pseudo multispectral classification (PMC) technique using CIE LAB spaces instead of gray scale T1-weighted MPRAGE images, combined with a new preprocessing technique for contrast enhancement and an optimized iterative K-means clustering. To improve the accuracy of the classification process, gray scale images were converted to multispectral CIE LAB data by applying several transformation matrices. Thus, the amount of information associated with each image voxel was increased. The image contrast was then enhanced by applying a real time function that separates brain tissue distributions and improve image contrast in certain brain regions. The data were then classified using an optimized iterative and convergent K-means classifier. The performance of the proposed approach was assessed using simulation and in vivo human studies through comparison with three common software packages used for brain MR image segmentation, namely FSL, SPM8 and K-means clustering. In the presence of high SNR, the results showed that the four algorithms achieve a good classification. Conversely, in the presence of low SNR, PMC was shown to outperform the other methods by accurately recovering all tissue volumes. The quantitative assessment of brain tissue classification for simulated studies showed that the PMC algorithm resulted in a mean Jaccard index (JI) of 0.74 compared to 0.75 for FSL, 0.7 for SPM and 0.8 for K-means. The in vivo human studies showed that the PMC algorithm resulted in a mean JI of 0.92, which reflects a good spatial overlap between segmented and actual volumes, compared to 0.84 for FSL, 0.78 for SPM and 0.66 for K-means. The proposed algorithm presents a high potential for improving the accuracy of automatic brain tissues classification and was found to be accurate even in the presence of high noise level.

Reproducibility and Bias in Healthy Brain Segmentation: Comparison of Two Popular Neuroimaging Platforms

We evaluated and compared the performance of two popular neuroimaging processing platforms: Statistical Parametric Mapping (SPM) and FMRIB Software Library (FSL). We focused on comparing brain segmentations using Kirby21, a magnetic resonance imaging (MRI) replication study with 21 subjects and two scans per subject conducted only a few hours apart. We tested within- and between-platform segmentation reliability both at the whole brain and in 10 regions of interest (ROIs). For a range of fixed probability thresholds we found no differences between-scans within-platform, but large differences between-platforms. We have also found very large differences between- and within-platforms when probability thresholds were changed. A randomized blinded reader study indicated that: (1) SPM and FSL performed well in terms of gray matter segmentation; (2) SPM and FSL performed poorly in terms of white matter segmentation; and (3) FSL slightly outperformed SPM in terms of CSF segmentation. We also found that tissue class probability thresholds can have profound effects on segmentation results. We conclude that the reproducibility of neuroimaging studies depends on the neuroimaging software-processing platform and tissue probability thresholds. Our results suggest that probability thresholds may not be comparable across platforms and consistency of results may be improved by estimating a probability threshold correspondence function between SPM and FSL.

Inter-Method Discrepancies in Brain Volume Estimation May Drive Inconsistent Findings in Autism

Previous studies applying automatic preprocessing methods on Structural Magnetic Resonance Imaging (sMRI) report inconsistent neuroanatomical abnormalities in Autism Spectrum Disorder (ASD). In this study we investigate inter-method differences as a possible cause behind these inconsistent findings. In particular, we focus on the estimation of the following brain volumes: gray matter (GM), white matter (WM), cerebrospinal fluid (CSF), and total intra cranial volume (TIV). T1-weighted sMRIs of 417 ASD subjects and 459 typically developing controls (TDC) from the ABIDE dataset were estimated using three popular preprocessing methods: SPM, FSL, and FreeSurfer (FS). Brain volumes estimated by the three methods were correlated but had significant inter-method differences; except TIVSPM vs. TIVFS, all inter-method differences were significant. ASD vs. TDC group differences in all brain volume estimates were dependent on the method used. SPM showed that TIV, GM, and CSF volumes of ASD were larger than TDC with statistical significance, whereas FS and FSL did not show significant differences in any of the volumes; in some cases, the direction of the differences were opposite to SPM. When methods were compared with each other, they showed differential biases for autism, and several biases were larger than ASD vs. TDC differences of the respective methods. After manual inspection, we found inter-method segmentation mismatches in the cerebellum, sub-cortical structures, and inter-sulcal CSF. In addition, to validate automated TIV estimates we performed manual segmentation on a subset of subjects. Results indicate that SPM estimates are closest to manual segmentation, followed by FS while FSL estimates were significantly lower. In summary, we show that ASD vs. TDC brain volume differences are method dependent and that these inter-method discrepancies can contribute to inconsistent neuroimaging findings in general. We suggest cross-validation across methods and emphasize the need to develop better methods to increase the robustness of neuroimaging findings.

Robust Intensity Standardization in Brain Magnetic Resonance Images

The paper is focused on a tiSsue-Based Standardization Technique (SBST) of magnetic resonance (MR) brain images. Magnetic Resonance Imaging intensities have no fixed tissue-specific numeric meaning, even within the same MRI protocol, for the same body region, or even for images of the same patient obtained on the same scanner in different moments. This affects postprocessing tasks such as automatic segmentation or unsupervised/supervised classification methods, which strictly depend on the observed image intensities, compromising the accuracy and efficiency of many image analyses algorithms. A large number of MR images from public databases, belonging to healthy people and to patients with different degrees of neurodegenerative pathology, were employed together with synthetic MRIs. Combining both histogram and tissue-specific intensity information, a correspondence is obtained for each tissue across images. The novelty consists of computing three standardizing transformations for the three main brain tissues, for each tissue class separately. In order to create a continuous intensity mapping, spline smoothing of the overall slightly discontinuous piecewise-linear intensity transformation is performed. The robustness of the technique is assessed in a post hoc manner, by verifying that automatic segmentation of images before and after standardization gives a high overlapping (Dice index >0.9) for each tissue class, even across images coming from different sources. Furthermore, SBST efficacy is tested by evaluating if and how much it increases intertissue discrimination and by assessing gaussianity of tissue gray-level distributions before and after standardization. Some quantitative comparisons to already existing different approaches available in the literature are performed.

Multi-atlas learner fusion: An efficient segmentation approach for large-scale data

We propose multi-atlas learner fusion (MLF), a framework for rapidly and accurately replicating the highly accurate, yet computationally expensive, multi-atlas segmentation framework based on fusing local learners. In the largest whole-brain multi-atlas study yet reported, multi-atlas segmentations are estimated for a training set of 3464 MR brain images. Using these multi-atlas estimates we (1) estimate a low-dimensional representation for selecting locally appropriate example images, and (2) build AdaBoost learners that map a weak initial segmentation to the multi-atlas segmentation result. Thus, to segment a new target image we project the image into the low-dimensional space, construct a weak initial segmentation, and fuse the trained, locally selected, learners. The MLF framework cuts the runtime on a modern computer from 36 h down to 3-8 min - a 270× speedup - by completely bypassing the need for deformable atlas-target registrations. Additionally, we (1) describe a technique for optimizing the weak initial segmentation and the AdaBoost learning parameters, (2) quantify the ability to replicate the multi-atlas result with mean accuracies approaching the multi-atlas intra-subject reproducibility on a testing set of 380 images, (3) demonstrate significant increases in the reproducibility of intra-subject segmentations when compared to a state-of-the-art multi-atlas framework on a separate reproducibility dataset, (4) show that under the MLF framework the large-scale data model significantly improve the segmentation over the small-scale model under the MLF framework, and (5) indicate that the MLF framework has comparable performance as state-of-the-art multi-atlas segmentation algorithms without using non-local information.

Feasibility of simultaneous whole-brain imaging on an integrated PET-MRI system using an enhanced 2-point Dixon attenuation correction method

PURPOSE

To evaluate a potential approach for improved attenuation correction (AC) of PET in simultaneous PET and MRI brain imaging, a straightforward approach that adds bone information missing on Dixon AC was explored.

METHODS

Bone information derived from individual T1-weighted MRI data using segmentation tools in SPM8, were added to the standard Dixon AC map. Percent relative difference between PET reconstructed with Dixon+bone and with Dixon AC maps were compared across brain regions of 13 oncology patients. The clinical potential of the improved Dixon AC was investigated by comparing relative perfusion (rCBF) measured with arterial spin labeling to relative glucose uptake (rPETdxbone) measured simultaneously with (18)F-flurodexoyglucose in several regions across the brain.

RESULTS

A gradual increase in PET signal from center to the edge of the brain was observed in PET reconstructed with Dixon+bone. A 5-20% reduction in regional PET signals were observed in data corrected with standard Dixon AC maps. These regional underestimations of PET were either reduced or removed when Dixon+bone AC was applied. The mean relative correlation coefficient between rCBF and rPETdxbone was r = 0.53 (p < 0.001). Marked regional variations in rCBF-to-rPET correlation were observed, with the highest associations in the caudate and cingulate and the lowest in limbic structures. All findings were well matched to observations from previous studies conducted with PET data reconstructed with computed tomography derived AC maps.

CONCLUSION

Adding bone information derived from T1-weighted MRI to Dixon AC maps can improve underestimation of PET activity in hybrid PET-MRI neuroimaging.

Self-assessed performance improves statistical fusion of image labels

PURPOSE

Expert manual labeling is the gold standard for image segmentation, but this process is difficult, time-consuming, and prone to inter-individual differences. While fully automated methods have successfully targeted many anatomies, automated methods have not yet been developed for numerous essential structures (e.g., the internal structure of the spinal cord as seen on magnetic resonance imaging). Collaborative labeling is a new paradigm that offers a robust alternative that may realize both the throughput of automation and the guidance of experts. Yet, distributing manual labeling expertise across individuals and sites introduces potential human factors concerns (e.g., training, software usability) and statistical considerations (e.g., fusion of information, assessment of confidence, bias) that must be further explored. During the labeling process, it is simple to ask raters to self-assess the confidence of their labels, but this is rarely done and has not been previously quantitatively studied. Herein, the authors explore the utility of self-assessment in relation to automated assessment of rater performance in the context of statistical fusion.

METHODS

The authors conducted a study of 66 volumes manually labeled by 75 minimally trained human raters recruited from the university undergraduate population. Raters were given 15 min of training during which they were shown examples of correct segmentation, and the online segmentation tool was demonstrated. The volumes were labeled 2D slice-wise, and the slices were unordered. A self-assessed quality metric was produced by raters for each slice by marking a confidence bar superimposed on the slice. Volumes produced by both voting and statistical fusion algorithms were compared against a set of expert segmentations of the same volumes.

RESULTS

Labels for 8825 distinct slices were obtained. Simple majority voting resulted in statistically poorer performance than voting weighted by self-assessed performance. Statistical fusion resulted in statistically indistinguishable performance from self-assessed weighted voting. The authors developed a new theoretical basis for using self-assessed performance in the framework of statistical fusion and demonstrated that the combined sources of information (both statistical assessment and self-assessment) yielded statistically significant improvement over the methods considered separately.

CONCLUSIONS

The authors present the first systematic characterization of self-assessed performance in manual labeling. The authors demonstrate that self-assessment and statistical fusion yield similar, but complementary, benefits for label fusion. Finally, the authors present a new theoretical basis for combining self-assessments with statistical label fusion.

Comparative evaluation of registration algorithms in different brain databases with varying difficulty: results and insights

Evaluating various algorithms for the inter-subject registration of brain magnetic resonance images (MRI) is a necessary topic receiving growing attention. Existing studies evaluated image registration algorithms in specific tasks or using specific databases (e.g., only for skull-stripped images, only for single-site images, etc.). Consequently, the choice of registration algorithms seems task- and usage/parameter-dependent. Nevertheless, recent large-scale, often multi-institutional imaging-related studies create the need and raise the question whether some registration algorithms can 1) generally apply to various tasks/databases posing various challenges; 2) perform consistently well, and while doing so, 3) require minimal or ideally no parameter tuning. In seeking answers to this question, we evaluated 12 general-purpose registration algorithms, for their generality, accuracy and robustness. We fixed their parameters at values suggested by algorithm developers as reported in the literature. We tested them in 7 databases/tasks, which present one or more of 4 commonly-encountered challenges: 1) inter-subject anatomical variability in skull-stripped images; 2) intensity homogeneity, noise and large structural differences in raw images; 3) imaging protocol and field-of-view (FOV) differences in multi-site data; and 4) missing correspondences in pathology-bearing images. Totally 7,562 registrations were performed. Registration accuracies were measured by (multi-)expert-annotated landmarks or regions of interest (ROIs). To ensure reproducibility, we used public software tools, public databases (whenever possible), and we fully disclose the parameter settings. We show evaluation results, and discuss the performances in light of algorithms' similarity metrics, transformation models and optimization strategies. We also discuss future directions for the algorithm development and evaluations.

Comparison of 10 brain tissue segmentation methods using revisited IBSR annotations

PURPOSE

Ground-truth annotations from the well-known Internet Brain Segmentation Repository (IBSR) datasets consider Sulcal cerebrospinal fluid (SCSF) voxels as gray matter. This can lead to bias when evaluating the performance of tissue segmentation methods. In this work we compare the accuracy of 10 brain tissue segmentation methods analyzing the effects of SCSF ground-truth voxels on accuracy estimations.

MATERIALS AND METHODS

The set of methods is composed by FAST, SPM5, SPM8, GAMIXTURE, ANN, FCM, KNN, SVPASEG, FANTASM, and PVC. Methods are evaluated using original IBSR ground-truth and ranked by means of their performance on pairwise comparisons using permutation tests. Afterward, the evaluation is repeated using IBSR ground-truth without considering SCSF.

RESULTS

The Dice coefficient of all methods is affected by changes in SCSF annotations, especially on SPM5, SPM8 and FAST. When not considering SCSF voxels, SVPASEG (0.90 ± 0.01) and SPM8 (0.91 ± 0.01) are the methods from our study that appear more suitable for gray matter tissue segmentation, while FAST (0.89 ± 0.02) is the best tool for segmenting white matter tissue.

CONCLUSION

The performance and the accuracy of methods on IBSR images vary notably when not considering SCSF voxels. The fact that three of the most common methods (FAST, SPM5, and SPM8) report an important change in their accuracy suggest to consider these differences in labeling for new comparative studies.

Investigation of brain tissue segmentation error and its effect on EEG source localization

Surgical therapy has become an important therapeutic alternative for patients with medically intractable epilepsy. Correct and anatomically precise localization of the epileptic focus, preferably with non-invasive methods, is the main goal of the pre-surgical epilepsy diagnosis to decide if resection of brain tissue is possible. For evaluating the performance of the source localization algorithms in an actual clinical situation, realistic patient-specific human head models that incorporate the heterogeneity nature of brain tissues is required. In this paper, performance of two of the most widely used software packages for brain segmentation, namely FSL and FreeSurfer has been analyzed. Then a segmented head model from a package with better performance is used to investigate the effects of brain tissue segmentation in EEG source localization.

Frontotemporoparietal asymmetry and lack of illness awareness in schizophrenia

INTRODUCTION

Lack of illness awareness or anosognosia occurs in both schizophrenia and right hemisphere lesions due to stroke, dementia, and traumatic brain injury. In the latter conditions, anosognosia is thought to arise from unilateral hemispheric dysfunction or interhemispheric disequilibrium, which provides an anatomical model for exploring illness unawareness in other neuropsychiatric disorders, such as schizophrenia.

METHODS

Both voxel-based morphometry using Diffeomorphic Anatomical Registration through Exponentiated Lie Algebra (DARTEL) and a deformation-based morphology analysis of hemispheric asymmetry were performed on 52 treated schizophrenia subjects, exploring the relationship between illness awareness and gray matter volume. Analyses included age, gender, and total intracranial volume as covariates.

RESULTS

Hemispheric asymmetry analyses revealed illness unawareness was significantly associated with right < left hemisphere volumes in the anteroinferior temporal lobe (t = 4.83, P = 0.051) using DARTEL, and the dorsolateral prefrontal cortex (t = 5.80, P = 0.003) and parietal lobe (t = 4.3, P = 0.050) using the deformation-based approach. Trend level associations were identified in the right medial prefrontal cortex (t = 4.49, P = 0.127) using DARTEL. Lack of illness awareness was also strongly associated with reduced total white matter volume (r = 0.401, P < 0.01) and illness severity (r = 0.559, P < 0.01).

CONCLUSION

These results suggest a relationship between anosognosia and hemispheric asymmetry in schizophrenia, supporting previous volume-based MRI studies in schizophrenia that found a relationship between illness unawareness and reduced right hemisphere gray matter volume. Functional imaging studies are required to examine the neural mechanisms contributing to these structural observations.

Non-local statistical label fusion for multi-atlas segmentation

Multi-atlas segmentation provides a general purpose, fully-automated approach for transferring spatial information from an existing dataset ("atlases") to a previously unseen context ("target") through image registration. The method to resolve voxelwise label conflicts between the registered atlases ("label fusion") has a substantial impact on segmentation quality. Ideally, statistical fusion algorithms (e.g., STAPLE) would result in accurate segmentations as they provide a framework to elegantly integrate models of rater performance. The accuracy of statistical fusion hinges upon accurately modeling the underlying process of how raters err. Despite success on human raters, current approaches inaccurately model multi-atlas behavior as they fail to seamlessly incorporate exogenous intensity information into the estimation process. As a result, locally weighted voting algorithms represent the de facto standard fusion approach in clinical applications. Moreover, regardless of the approach, fusion algorithms are generally dependent upon large atlas sets and highly accurate registration as they implicitly assume that the registered atlases form a collectively unbiased representation of the target. Herein, we propose a novel statistical fusion algorithm, Non-Local STAPLE (NLS). NLS reformulates the STAPLE framework from a non-local means perspective in order to learn what label an atlas would have observed, given perfect correspondence. Through this reformulation, NLS (1) seamlessly integrates intensity into the estimation process, (2) provides a theoretically consistent model of multi-atlas observation error, and (3) largely diminishes the need for large atlas sets and very high-quality registrations. We assess the sensitivity and optimality of the approach and demonstrate significant improvement in two empirical multi-atlas experiments.

Formulating spatially varying performance in the statistical fusion framework

To date, label fusion methods have primarily relied either on global [e.g., simultaneous truth and performance level estimation (STAPLE), globally weighted vote] or voxelwise (e.g., locally weighted vote) performance models. Optimality of the statistical fusion framework hinges upon the validity of the stochastic model of how a rater errs (i.e., the labeling process model). Hitherto, approaches have tended to focus on the extremes of potential models. Herein, we propose an extension to the STAPLE approach to seamlessly account for spatially varying performance by extending the performance level parameters to account for a smooth, voxelwise performance level field that is unique to each rater. This approach, Spatial STAPLE, provides significant improvements over state-of-the-art label fusion algorithms in both simulated and empirical data sets.

Impact of Markov Random Field optimizer on MRI-based tissue segmentation in the aging brain

Automatically segmenting brain magnetic resonance images into grey matter, white matter, and cerebrospinal fluid compartments is a fundamentally important neuroimaging problem whose difficulty is heightened in the presence of aging and neurodegenerative disease. Current methods overlap greatly in terms of identifiable algorithmic components, and the impact of specific components on performance is generally unclear in important real-world scenarios involving serial scanning, multiple scanners, and neurodegenerative disease. Therefore we evaluated the impact that one such component, the Markov Random Field (MRF) optimizer that encourages spatially-smooth tissue labelings, has on brain tissue segmentation performance. Two challenging elderly data sets were used to test segmentation consistency across scanners and biological plausibility of tissue change estimates; and a simulated young brain data set was used to test accuracy against ground truth. Belief propagation (BP) and graph cuts (GC), used as the MRF optimizer component of a standardized segmentation system, provide high segmentation performance on aggregate that is competitive with end-to-end systems provided by SPM and FSL (FAST) as well as the more traditional MRF optimizer iterated conditional modes (ICM). However, the relative performance of each method varied strongly by performance criterion and differed between young and old brains. The findings emphasize the unique difficulties involved in segmenting the aging brain, and suggest that optimal algorithm components may depend in part on performance criteria.

Foibles, follies, and fusion: web-based collaboration for medical image labeling

Labels that identify specific anatomical and functional structures within medical images are essential to the characterization of the relationship between structure and function in many scientific and clinical studies. Automated methods that allow for high throughput have not yet been developed for all anatomical targets or validated for exceptional anatomies, and manual labeling remains the gold standard in many cases. However, manual placement of labels within a large image volume such as that obtained using magnetic resonance imaging (MRI) is exceptionally challenging, resource intensive, and fraught with intra- and inter-rater variability. The use of statistical methods to combine labels produced by multiple raters has grown significantly in popularity, in part, because it is thought that by estimating and accounting for rater reliability estimates of the true labels will be more accurate. This paper demonstrates the performance of a class of these statistical label combination methodologies using real-world data contributed by minimally trained human raters. The consistency of the statistical estimates, the accuracy compared to the individual observations, and the variability of both the estimates and the individual observations with respect to the number of labels are presented. It is demonstrated that statistical fusion successfully combines label information using data from online (Internet-based) collaborations among minimally trained raters. This first successful demonstration of a statistically based approach using minimally trained raters opens numerous possibilities for very large scale efforts in collaboration. Extension and generalization of these technologies for new applications will certainly present fascinating areas for continuing research.

Simple paradigm for extra-cerebral tissue removal: algorithm and analysis

Extraction of the brain-i.e. cerebrum, cerebellum, and brain stem-from T1-weighted structural magnetic resonance images is an important initial step in neuroimage analysis. Although automatic algorithms are available, their inconsistent handling of the cortical mantle often requires manual interaction, thereby reducing their effectiveness. This paper presents a fully automated brain extraction algorithm that incorporates elastic registration, tissue segmentation, and morphological techniques which are combined by a watershed principle, while paying special attention to the preservation of the boundary between the gray matter and the cerebrospinal fluid. The approach was evaluated by comparison to a manual rater, and compared to several other leading algorithms on a publically available data set of brain images using the Dice coefficient and containment index as performance metrics. The qualitative and quantitative impact of this initial step on subsequent cortical surface generation is also presented. Our experiments demonstrate that our approach is quantitatively better than six other leading algorithms (with statistical significance on modern T1-weighted MR data). We also validated the robustness of the algorithm on a very large data set of over one thousand subjects, and showed that it can replace an experienced manual rater as preprocessing for a cortical surface extraction algorithm with statistically insignificant differences in cortical surface position.

Brain MRI tissue classification based on local Markov random fields

A new method for tissue classification of brain magnetic resonance images (MRI) of the brain is proposed. The method is based on local image models where each models the image content in a subset of the image domain. With this local modeling approach, the assumption that tissue types have the same characteristics over the brain needs not to be evoked. This is important because tissue type characteristics, such as T1 and T2 relaxation times and proton density, vary across the individual brain and the proposed method offers improved protection against intensity non-uniformity artifacts that can hamper automatic tissue classification methods in brain MRI. A framework in which local models for tissue intensities and Markov Random Field (MRF) priors are combined into a global probabilistic image model is introduced. This global model will be an inhomogeneous MRF and it can be solved by standard algorithms such as iterative conditional modes. The division of the whole image domain into local brain regions possibly having different intensity statistics is realized via sub-volume probabilistic atlases. Finally, the parameters for the local intensity models are obtained without supervision by maximizing the weighted likelihood of a certain finite mixture model. For the maximization task, a novel genetic algorithm almost free of initialization dependency is applied. The algorithm is tested on both simulated and real brain MR images. The experiments confirm that the new method offers a useful improvement of the tissue classification accuracy when the basic tissue characteristics vary across the brain and the noise level of the images is reasonable. The method also offers better protection against intensity non-uniformity artifact than the corresponding method based on a global (whole image) modeling scheme.