Pattern classification of fMRI data: applications for analysis of spatially distributed cortical networks

Beta-band power classification of go/no-go arm-reaching responses in the human hippocampus

Objective.Can we classify movement execution and inhibition from hippocampal oscillations during arm-reaching tasks? Traditionally associated with memory encoding, spatial navigation, and motor sequence consolidation, the hippocampus has come under scrutiny for its potential role in movement processing. Stereotactic electroencephalography (SEEG) has provided a unique opportunity to study the neurophysiology of the human hippocampus during motor tasks. In this study, we assess the accuracy of discriminant functions, in combination with principal component analysis (PCA), in classifying between 'Go' and 'No-go' trials in a Go/No-go arm-reaching task.Approach.Our approach centers on capturing the modulation of beta-band (13-30 Hz) power from multiple SEEG contacts in the hippocampus and minimizing the dimensional complexity of channels and frequency bins. This study utilizes SEEG data from the human hippocampus of 10 participants diagnosed with epilepsy. Spectral power was computed during a 'center-out' Go/No-go arm-reaching task, where participants reached or withheld their hand based on a colored cue. PCA was used to reduce data dimension and isolate the highest-variance components within the beta band. The Silhouette score was employed to measure the quality of clustering between 'Go' and 'No-go' trials. The accuracy of five different discriminant functions was evaluated using cross-validation.Main results.The Diagonal-Quadratic model performed best of the 5 classification models, exhibiting the lowest error rate in all participants (median: 9.91%, average: 14.67%). PCA showed that the first two principal components collectively accounted for 54.83% of the total variance explained on average across all participants, ranging from 36.92% to 81.25% among participants.Significance.This study shows that PCA paired with a Diagonal-Quadratic model can be an effective method for classifying between Go/No-go trials from beta-band power in the hippocampus during arm-reaching responses. This emphasizes the significance of hippocampal beta-power modulation in motor control, unveiling its potential implications for brain-computer interface applications.

Tree representations of brain structural connectivity via persistent homology

The brain structural connectome is generated by a collection of white matter fiber bundles constructed from diffusion weighted MRI (dMRI), acting as highways for neural activity. There has been abundant interest in studying how the structural connectome varies across individuals in relation to their traits, ranging from age and gender to neuropsychiatric outcomes. After applying tractography to dMRI to get white matter fiber bundles, a key question is how to represent the brain connectome to facilitate statistical analyses relating connectomes to traits. The current standard divides the brain into regions of interest (ROIs), and then relies on an adjacency matrix (AM) representation. Each cell in the AM is a measure of connectivity, e.g., number of fiber curves, between a pair of ROIs. Although the AM representation is intuitive, a disadvantage is the high-dimensionality due to the large number of cells in the matrix. This article proposes a simpler tree representation of the brain connectome, which is motivated by ideas in computational topology and takes topological and biological information on the cortical surface into consideration. We demonstrate that our tree representation preserves useful information and interpretability, while reducing dimensionality to improve statistical and computational efficiency. Applications to data from the Human Connectome Project (HCP) are considered and code is provided for reproducing our analyses.

A Non-cognitive Behavioral Model for Interpreting Functional Neuroimaging Studies

The dominant model for interpreting brain imaging experiments, which we refer to as the Standard Cognitive Model (SCM), assumes that the brain is organized in support of mental processes that control behavior. However, functional neuroimaging experiments of cognitive tasks have not shown clear anatomic segregation between mental processes originally proposed by this model. This failing has been blamed on limitations in imaging technology and non-linearity in the brain's implementation of these processes. However, the validity of the underlying cognitive models used to describe the brain has rarely been questioned or directly tested against imaging results. We propose an alternative model of brain function, that we term the Non-cognitive Behavioral Model (NBM), which correlates observed human behavior directly with measured brain activity without making assumptions about intervening cognitive processes. Our model derives from behavioral psychology but is extended to include brain activity, in addition to behavior, as observables. A further extension is the role of neuroplasticity, as opposed to innate cognitive processes, in developing the brain's support of cognitive behavior. We present the theoretical basis with which the SCM maps cognitive processes onto functional magnetic resonance and positron emission tomography images and compare and contrast with the NBM. We also describe how the NBM can be used experimentally to study how the brain supports behavior. Two applications are presented that support the usefulness of the NBM. In one, the NBM use of the total functional imaging signal (not just the differences between states) provides a stronger correlation of neural activity with the behavioral state of consciousness than the SCM approach in both anesthesia and coma. The second example reviews studies of facial and object recognition that provide evidence for the NBM proposal that neuroplasticity and experience play key roles in the brain's support of recognition and other behaviors. The conclusions regarding neuroplasticity are then generalized to explain the incomplete functional segregation observed in the application of the SCM to neuroimaging.

Reprint of: Mapping human brain lesions and their functional consequences

Neuroscience has a long history of inferring brain function by examining the relationship between brain injury and subsequent behavioral impairments. The primary advantage of this method over correlative methods is that it can tell us if a certain brain region is necessary for a given cognitive function. In addition, lesion-based analyses provide unique insights into clinical deficits. In the last decade, statistical voxel-based lesion behavior mapping (VLBM) emerged as a powerful method for understanding the architecture of the human brain. This review illustrates how VLBM improves our knowledge of functional brain architecture, as well as how it is inherently limited by its mass-univariate approach. A wide array of recently developed methods appear to supplement traditional VLBM. This paper provides an overview of these new methods, including the use of specialized imaging modalities, the combination of structural imaging with normative connectome data, as well as multivariate analyses of structural imaging data. We see these new methods as complementing rather than replacing traditional VLBM, providing synergistic tools to answer related questions. Finally, we discuss the potential for these methods to become established in cognitive neuroscience and in clinical applications.

Neuroimaging Applications in Dystonia

Dystonia is a neurological disorder characterized by involuntary, repetitive movements. Although the precise mechanisms of dystonia development remain unknown, the diversity of its clinical phenotypes is thought to be associated with multifactorial pathophysiology, which is linked not only to alterations of brain organization, but also environmental stressors and gene mutations. This chapter will present an overview of the pathophysiology of isolated dystonia through the lens of applications of major neuroimaging methodologies, with links to genetics and environmental factors that play a prominent role in symptom manifestation.

Functional MRI of Handwriting Tasks: A Study of Healthy Young Adults Interacting with a Novel Touch-Sensitive Tablet

Handwriting is a complex human activity that engages a blend of cognitive and visual motor skills. Current understanding of the neural correlates of handwriting has largely come from lesion studies of patients with impaired handwriting. Task-based fMRI studies would be useful to supplement this work. To address concerns over ecological validity, previously we developed a fMRI-compatible, computerized tablet system for writing and drawing including visual feedback of hand position and an augmented reality display. The purpose of the present work is to use the tablet system in proof-of-concept to characterize brain activity associated with clinically relevant handwriting tasks, originally developed to characterize handwriting impairments in Alzheimer’s disease patients. As a prelude to undertaking fMRI studies of patients, imaging was performed of twelve young healthy subjects who copied sentences, phone numbers, and grocery lists using the fMRI-compatible tablet. Activation maps for all handwriting tasks consisted of a distributed network of regions in reasonable agreement with previous studies of handwriting performance. In addition, differences in brain activity were observed between the test subcomponents consistent with different demands of neural processing for successful task performance, as identified by investigating three quantitative behavioral metrics (writing speed, stylus contact force and stylus in air time). This study provides baseline behavioral and brain activity results for fMRI studies that adopt this handwriting test to characterize patients with brain impairments.

Mapping human brain lesions and their functional consequences

Neuroscience has a long history of inferring brain function by examining the relationship between brain injury and subsequent behavioral impairments. The primary advantage of this method over correlative methods is that it can tell us if a certain brain region is necessary for a given cognitive function. In addition, lesion-based analyses provide unique insights into clinical deficits. In the last decade, statistical voxel-based lesion behavior mapping (VLBM) emerged as a powerful method for understanding the architecture of the human brain. This review illustrates how VLBM improves our knowledge of functional brain architecture, as well as how it is inherently limited by its mass-univariate approach. A wide array of recently developed methods appear to supplement traditional VLBM. This paper provides an overview of these new methods, including the use of specialized imaging modalities, the combination of structural imaging with normative connectome data, as well as multivariate analyses of structural imaging data. We see these new methods as complementing rather than replacing traditional VLBM, providing synergistic tools to answer related questions. Finally, we discuss the potential for these methods to become established in cognitive neuroscience and in clinical applications.

Tablet-Based Functional MRI of the Trail Making Test: Effect of Tablet Interaction Mode

The Trail Making Test (TMT) is widely used for assessing executive function, frontal lobe abilities, and visual motor skills. Part A of this pen-and-paper test (TMT-A) involves linking numbers randomly distributed in space, in ascending order. Part B (TMT-B) alternates between linking numbers and letters. TMT-B is more demanding than TMT-A, but the mental processing that supports the performance of this test remains incompletely understood. Functional MRI (fMRI) may help to clarify the relationship between TMT performance and brain activity, but providing an environment that supports real-world pen-and-paper interactions during fMRI is challenging. Previously, an fMRI-compatible tablet system was developed for writing and drawing with two modes of interaction: the original cursor-based, proprioceptive approach, and a new mode involving augmented reality to provide visual feedback of hand position (VFHP) for enhanced user interaction. This study characterizes the use of the tablet during fMRI of young healthy adults (n = 22), with half of the subjects performing TMT with VFHP and the other half performing TMT without VFHP. Activation maps for both TMT-A and TMT-B performance showed considerable overlap between the two tablet modes, and no statistically differences in brain activity were detected when contrasting TMT-B vs. TMT-A for the two tablet modes. Behavioral results also showed no statistically different interaction effects for TMT-B vs. TMT-A for the two tablet modes. Tablet-based TMT scores showed reasonable convergent validity with those obtained by administering the standard pen-and-paper TMT to the same subjects. Overall, the results suggest that despite the slightly different mechanisms involved for the two modes of tablet interaction, both are suitable for use in fMRI studies involving TMT performance. This study provides information for using tablet-based TMT methods appropriately in future fMRI studies involving patients and healthy individuals.

Multivariate Connectome-Based Symptom Mapping in Post-Stroke Patients: Networks Supporting Language and Speech

Language processing relies on a widespread network of brain regions. Univariate post-stroke lesion-behavior mapping is a particularly potent method to study brain-language relationships. However, it is a concern that this method may overlook structural disconnections to seemingly spared regions and may fail to adjudicate between regions that subserve different processes but share the same vascular perfusion bed. For these reasons, more refined structural brain mapping techniques may improve the accuracy of detecting brain networks supporting language. In this study, we applied a predictive multivariate framework to investigate the relationship between language deficits in human participants with chronic aphasia and the topological distribution of structural brain damage, defined as post-stroke necrosis or cortical disconnection. We analyzed lesion maps as well as structural connectome measures of whole-brain neural network integrity to predict clinically applicable language scores from the Western Aphasia Battery (WAB). Out-of-sample prediction accuracy was comparable for both types of analyses, which revealed spatially distinct, albeit overlapping, networks of cortical regions implicated in specific aspects of speech functioning. Importantly, all WAB scores could be predicted at better-than-chance level from the connections between gray-matter regions spared by the lesion. Connectome-based analysis highlighted the role of connectivity of the temporoparietal junction as a multimodal area crucial for language tasks. Our results support that connectome-based approaches are an important complement to necrotic lesion-based approaches and should be used in combination with lesion mapping to fully elucidate whether structurally damaged or structurally disconnected regions relate to aphasic impairment and its recovery.

SIGNIFICANCE STATEMENT

We present a novel multivariate approach of predicting post-stroke impairment of speech and language from the integrity of the connectome. We compare it with multivariate prediction of speech and language scores from lesion maps, using cross-validation framework and a large (n = 90) database of behavioral and neuroimaging data from individuals with post-stroke aphasia. Connectome-based analysis was similar to lesion-based analysis in terms of predictive accuracy and provided additional details about the importance of specific connections (in particular, between parietal and posterior temporal areas) for preserving speech functions. Our results suggest that multivariate predictive analysis of the connectome is a useful complement to multivariate lesion analysis, being less dependent on the spatial constraints imposed by underlying vasculature.

Mapping complementary features of cross-species structural connectivity to construct realistic "Virtual Brains"

Modern systems neuroscience increasingly leans on large-scale multi-lab neuroinformatics initiatives to provide necessary capacity for biologically realistic modeling of primate whole-brain activity. Here, we present a framework to assemble primate brain's biologically plausible anatomical backbone for such modeling initiatives. In this framework, structural connectivity is determined by adding complementary information from invasive macaque axonal tract tracing and non-invasive human diffusion tensor imaging. Both modalities are combined by means of available interspecies registration tools and a newly developed Bayesian probabilistic modeling approach to extract common connectivity evidence. We demonstrate how this novel framework is embedded in the whole-brain simulation platform called The Virtual Brain (TVB). Hum Brain Mapp 38:2080-2093, 2017. © 2017 Wiley Periodicals, Inc.

Cortical sensorimotor alterations classify clinical phenotype and putative genotype of spasmodic dysphonia

BACKGROUND AND PURPOSE

Spasmodic dysphonia (SD), or laryngeal dystonia, is a task-specific isolated focal dystonia of unknown causes and pathophysiology. Although functional and structural abnormalities have been described in this disorder, the influence of its different clinical phenotypes and genotypes remains scant, making it difficult to explain SD pathophysiology and to identify potential biomarkers.

METHODS

We used a combination of independent component analysis and linear discriminant analysis of resting-state functional magnetic resonance imaging data to investigate brain organization in different SD phenotypes (abductor versus adductor type) and putative genotypes (familial versus sporadic cases) and to characterize neural markers for genotype/phenotype categorization.

RESULTS

We found abnormal functional connectivity within sensorimotor and frontoparietal networks in patients with SD compared with healthy individuals as well as phenotype- and genotype-distinct alterations of these networks, involving primary somatosensory, premotor and parietal cortices. The linear discriminant analysis achieved 71% accuracy classifying SD and healthy individuals using connectivity measures in the left inferior parietal and sensorimotor cortices. When categorizing between different forms of SD, the combination of measures from the left inferior parietal, premotor and right sensorimotor cortices achieved 81% discriminatory power between familial and sporadic SD cases, whereas the combination of measures from the right superior parietal, primary somatosensory and premotor cortices led to 71% accuracy in the classification of adductor and abductor SD forms.

CONCLUSIONS

Our findings present the first effort to identify and categorize isolated focal dystonia based on its brain functional connectivity profile, which may have a potential impact on the future development of biomarkers for this rare disorder.

Linear Discriminant Analysis Achieves High Classification Accuracy for the BOLD fMRI Response to Naturalistic Movie Stimuli

Naturalistic stimuli like movies evoke complex perceptual processes, which are of great interest in the study of human cognition by functional MRI (fMRI). However, conventional fMRI analysis based on statistical parametric mapping (SPM) and the general linear model (GLM) is hampered by a lack of accurate parametric models of the BOLD response to complex stimuli. In this situation, statistical machine-learning methods, a.k.a. multivariate pattern analysis (MVPA), have received growing attention for their ability to generate stimulus response models in a data-driven fashion. However, machine-learning methods typically require large amounts of training data as well as computational resources. In the past, this has largely limited their application to fMRI experiments involving small sets of stimulus categories and small regions of interest in the brain. By contrast, the present study compares several classification algorithms known as Nearest Neighbor (NN), Gaussian Naïve Bayes (GNB), and (regularized) Linear Discriminant Analysis (LDA) in terms of their classification accuracy in discriminating the global fMRI response patterns evoked by a large number of naturalistic visual stimuli presented as a movie. Results show that LDA regularized by principal component analysis (PCA) achieved high classification accuracies, above 90% on average for single fMRI volumes acquired 2 s apart during a 300 s movie (chance level 0.7% = 2 s/300 s). The largest source of classification errors were autocorrelations in the BOLD signal compounded by the similarity of consecutive stimuli. All classifiers performed best when given input features from a large region of interest comprising around 25% of the voxels that responded significantly to the visual stimulus. Consistent with this, the most informative principal components represented widespread distributions of co-activated brain regions that were similar between subjects and may represent functional networks. In light of these results, the combination of naturalistic movie stimuli and classification analysis in fMRI experiments may prove to be a sensitive tool for the assessment of changes in natural cognitive processes under experimental manipulation.

Predicting aphasia type from brain damage measured with structural MRI

Chronic aphasia is a common consequence of a left-hemisphere stroke. Since the early insights by Broca and Wernicke, studying the relationship between the loci of cortical damage and patterns of language impairment has been one of the concerns of aphasiology. We utilized multivariate classification in a cross-validation framework to predict the type of chronic aphasia from the spatial pattern of brain damage. Our sample consisted of 98 patients with five types of aphasia (Broca's, Wernicke's, global, conduction, and anomic), classified based on scores on the Western Aphasia Battery (WAB). Binary lesion maps were obtained from structural MRI scans (obtained at least 6 months poststroke, and within 2 days of behavioural assessment); after spatial normalization, the lesions were parcellated into a disjoint set of brain areas. The proportion of damage to the brain areas was used to classify patients' aphasia type. To create this parcellation, we relied on five brain atlases; our classifier (support vector machine - SVM) could differentiate between different kinds of aphasia using any of the five parcellations. In our sample, the best classification accuracy was obtained when using a novel parcellation that combined two previously published brain atlases, with the first atlas providing the segmentation of grey matter, and the second atlas used to segment the white matter. For each aphasia type, we computed the relative importance of different brain areas for distinguishing it from other aphasia types; our findings were consistent with previously published reports of lesion locations implicated in different types of aphasia. Overall, our results revealed that automated multivariate classification could distinguish between aphasia types based on damage to atlas-defined brain areas.

Relevant feature set estimation with a knock-out strategy and random forests

Group analysis of neuroimaging data is a vital tool for identifying anatomical and functional variations related to diseases as well as normal biological processes. The analyses are often performed on a large number of highly correlated measurements using a relatively smaller number of samples. Despite the correlation structure, the most widely used approach is to analyze the data using univariate methods followed by post-hoc corrections that try to account for the data's multivariate nature. Although widely used, this approach may fail to recover from the adverse effects of the initial analysis when local effects are not strong. Multivariate pattern analysis (MVPA) is a powerful alternative to the univariate approach for identifying relevant variations. Jointly analyzing all the measures, MVPA techniques can detect global effects even when individual local effects are too weak to detect with univariate analysis. Current approaches are successful in identifying variations that yield highly predictive and compact models. However, they suffer from lessened sensitivity and instabilities in identification of relevant variations. Furthermore, current methods' user-defined parameters are often unintuitive and difficult to determine. In this article, we propose a novel MVPA method for group analysis of high-dimensional data that overcomes the drawbacks of the current techniques. Our approach explicitly aims to identify all relevant variations using a "knock-out" strategy and the Random Forest algorithm. In evaluations with synthetic datasets the proposed method achieved substantially higher sensitivity and accuracy than the state-of-the-art MVPA methods, and outperformed the univariate approach when the effect size is low. In experiments with real datasets the proposed method identified regions beyond the univariate approach, while other MVPA methods failed to replicate the univariate results. More importantly, in a reproducibility study with the well-known ADNI dataset the proposed method yielded higher stability and power than the univariate approach.

Discovering networks altered by potential threat ("anxiety") using quadratic discriminant analysis

Researchers have only recently begun using functional neuroimaging to explore the human response to periods of sustained anxious anticipation, namely potential threat. Here, we investigated brain responses acquired with functional MRI during an instructed threat of shock paradigm used to create sustained periods of aversive anticipation. In this re-analysis of previously published data, we employed quadratic discriminant analysis to classify the multivariate pattern of whole-brain functional connectivity and to identify connectivity changes during periods of potential threat. Our method identifies clusters with altered connectivity on a voxelwise basis, thus eschewing the need to define regions a priori. Classifier generalization was evaluated by testing on data from participants not used during training. Robust classification between threat and safe contexts was possible, and inspection of "diagnostic features" revealed altered functional connectivity involving the intraparietal sulcus, task-negative regions, striatum, and anterior cingulate cortex. We anticipate that the proposed method will prove useful to experimenters wishing to identify large-scale functional networks that distinguish between experimental conditions or groups.

Is the preference of natural versus man-made scenes driven by bottom-up processing of the visual features of nature?

Previous research has shown that viewing images of nature scenes can have a beneficial effect on memory, attention, and mood. In this study, we aimed to determine whether the preference of natural versus man-made scenes is driven by bottom-up processing of the low-level visual features of nature. We used participants' ratings of perceived naturalness as well as esthetic preference for 307 images with varied natural and urban content. We then quantified 10 low-level image features for each image (a combination of spatial and color properties). These features were used to predict esthetic preference in the images, as well as to decompose perceived naturalness to its predictable (modeled by the low-level visual features) and non-modeled aspects. Interactions of these separate aspects of naturalness with the time it took to make a preference judgment showed that naturalness based on low-level features related more to preference when the judgment was faster (bottom-up). On the other hand, perceived naturalness that was not modeled by low-level features was related more to preference when the judgment was slower. A quadratic discriminant classification analysis showed how relevant each aspect of naturalness (modeled and non-modeled) was to predicting preference ratings, as well as the image features on their own. Finally, we compared the effect of color-related and structure-related modeled naturalness, and the remaining unmodeled naturalness in predicting esthetic preference. In summary, bottom-up (color and spatial) properties of natural images captured by our features and the non-modeled naturalness are important to esthetic judgments of natural and man-made scenes, with each predicting unique variance.

The perception of naturalness correlates with low-level visual features of environmental scenes

Previous research has shown that interacting with natural environments vs. more urban or built environments can have salubrious psychological effects, such as improvements in attention and memory. Even viewing pictures of nature vs. pictures of built environments can produce similar effects. A major question is: What is it about natural environments that produces these benefits? Problematically, there are many differing qualities between natural and urban environments, making it difficult to narrow down the dimensions of nature that may lead to these benefits. In this study, we set out to uncover visual features that related to individuals' perceptions of naturalness in images. We quantified naturalness in two ways: first, implicitly using a multidimensional scaling analysis and second, explicitly with direct naturalness ratings. Features that seemed most related to perceptions of naturalness were related to the density of contrast changes in the scene, the density of straight lines in the scene, the average color saturation in the scene and the average hue diversity in the scene. We then trained a machine-learning algorithm to predict whether a scene was perceived as being natural or not based on these low-level visual features and we could do so with 81% accuracy. As such we were able to reliably predict subjective perceptions of naturalness with objective low-level visual features. Our results can be used in future studies to determine if these features, which are related to naturalness, may also lead to the benefits attained from interacting with nature.