Learning Sparse Networks and Brain Health Biomarkers from fMRI Using Dictionary Learning Methods

THINGS-data reflects three large-scale neuroimaging and behavioral datasets of object processing in humans, comprising densely sampled functional MRI and magnetoencephalographic recordings, as well as 4.70 million similarity judgments in response to thousands of photographic images for up to 1854 objects.

Automatic prostate segmentation from MR images is an important task in various clinical applications such as prostate cancer staging and MR-guided radiotherapy planning. However, the large appearance and shape variations of the prostate in MR images make the segmentation problem difficult to solve. Traditional Active Shape/Appearance Model (ASM/AAM) has limited accuracy on this problem, since its basic assumption, i.e., both shape and appearance of the targeted organ follow Gaussian distributions, is invalid in prostate MR images. To this end, the authors propose a sparse dictionary learning method to model the image appearance in a nonparametric fashion and further integrate the appearance model into a deformable segmentation framework for prostate MR segmentation. To drive the deformable model for prostate segmentation, the authors propose nonparametric appearance and shape models. The nonparametric appearance model is based on a novel dictionary learning method, namely distributed discriminative dictionary (DDD) learning, which is able to capture fine distinctions in image appearance. To increase the differential power of traditional dictionary-based classification methods, the authors' DDD learning approach takes three strategies. First, two dictionaries for prostate and nonprostate tissues are built, respectively, using the discriminative features obtained from minimum redundancy maximum relevance feature selection. Second, linear discriminant analysis is employed as a linear classifier to boost the optimal separation between prostate and nonprostate tissues, based on the representation residuals from sparse representation. Third, to enhance the robustness of the authors' classification method, multiple local dictionaries are learned for local regions along the prostate boundary (each with small appearance variations), instead of learning one global classifier for the entire prostate. These discriminative dictionaries are located on different patches of the prostate surface and trained to adaptively capture the appearance in different prostate zones, thus achieving better local tissue differentiation. For each local region, multiple classifiers are trained based on the randomly selected samples and finally assembled by a specific fusion method. In addition to this nonparametric appearance model, a prostate shape model is learned from the shape statistics using a novel approach, sparse shape composition, which can model nonGaussian distributions of shape variation and regularize the 3D mesh deformation by constraining it within the observed shape subspace. The proposed method has been evaluated on two datasets consisting of T2-weighted MR prostate images. For the first (internal) dataset, the classification effectiveness of the authors' improved dictionary learning has been validated by comparing it with three other variants of traditional dictionary learning methods. The experimental results show that the authors' method yields a Dice Ratio of 89.1% compared to the manual segmentation, which is more accurate than the three state-of-the-art MR prostate segmentation methods under comparison. For the second dataset, the MICCAI 2012 challenge dataset, the authors' proposed method yields a Dice Ratio of 87.4%, which also achieves better segmentation accuracy than other methods under comparison. A new magnetic resonance image prostate segmentation method is proposed based on the combination of deformable model and dictionary learning methods, which achieves more accurate segmentation performance on prostate T2 MR images.

Analysis of time-evolving data is crucial to understand the functioning of dynamic systems such as the brain. For instance, analysis of functional magnetic resonance imaging (fMRI) data collected during a task may reveal spatial regions of interest, and how they evolve during the task. However, capturing underlying spatial patterns as well as their change in time is challenging. The traditional approach in fMRI data analysis is to assume that underlying spatial regions of interest are static. In this article, using fractional amplitude of low-frequency fluctuations (fALFF) as an effective way to summarize the variability in fMRI data collected during a task, we arrange time-evolving fMRI data as a subjects by voxels by time windows tensor, and analyze the tensor using a tensor factorization-based approach called a PARAFAC2 model to reveal spatial dynamics. The PARAFAC2 model jointly analyzes data from multiple time windows revealing subject-mode patterns, evolving spatial regions (also referred to as networks) and temporal patterns. We compare the PARAFAC2 model with matrix factorization-based approaches relying on independent components, namely, joint independent component analysis (ICA) and independent vector analysis (IVA), commonly used in neuroimaging data analysis. We assess the performance of the methods in terms of capturing evolving networks through extensive numerical experiments demonstrating their modeling assumptions. In particular, we show that (i) PARAFAC2 provides a compact representation in all modes, i.e., subjects, time, and voxels, revealing temporal patterns as well as evolving spatial networks, (ii) joint ICA is as effective as PARAFAC2 in terms of revealing evolving networks but does not reveal temporal patterns, (iii) IVA's performance depends on sample size, data distribution and covariance structure of underlying networks. When these assumptions are satisfied, IVA is as accurate as the other methods, (iv) when subject-mode patterns differ from one time window to another, IVA is the most accurate. Furthermore, we analyze real fMRI data collected during a sensory motor task, and demonstrate that a component indicating statistically significant group difference between patients with schizophrenia and healthy controls is captured, which includes primary and secondary motor regions, cerebellum, and temporal lobe, revealing a meaningful spatial map and its temporal change.

Fuzzy cluster analysis (FCA) is a new exploratory method for analyzing fMRI data. Using simulated functional MRI (fMRI) data, the performance of FCA, as implemented in the software package Evident, was tested and a quantitative comparison with correlation analysis is presented. Furthermore, the fMRI model fit allows separation and quantification of flow and blood oxygen level dependent (BOLD) contributions in the human visual cortex. In gradient-recalled echo fMRI at 1.5 T (TR = 60 ms, TE = 42 ms, radiofrequency excitation flip angle [theta] = 10 degrees-60 degrees) total signal enhancement in the human visual cortex, ie, flow-enhanced BOLD plus inflow contributions, on average varies from 5% to 10% in or close to the visual cortex (average cerebral blood volume [CBV] = 4%) and from 100% to 20% in areas containing medium-sized vessels (ie, average CBV = 12% per voxel), respectively. Inflow enhancement, however, is restricted to intravascular space (= CBV) and increases with increasing radiofrequency (RF) flip angle, whereas BOLD contributions may be obtained from a region up to three times larger and, applying an unspoiled gradient-echo (GRE) sequence, also show a flip angle dependency with a minimum at approximately 30 degrees. This result suggests that a localized hemodynamic response from the microvasculature at 1.5 T may be extracted via fuzzy clustering. In summary, fuzzy clustering of fMRI data, as realized in the Evident software, is a robust and efficient method to (a) separate functional brain activation from noise or other sources resulting in time-dependent signal changes as proven by simulated fMRI data analysis and in vivo data from the visual cortex, and (b) allows separation of different levels of activation even if the temporal pattern is indistinguishable. Combining fuzzy cluster separation of brain activation with appropriate model calculations allows quantification of flow and (flow-enhanced) BOLD contributions in areas with different vascularization.

Default-mode network (DMN) activity measured with functional magnetic resonance imaging (fMRI) represents dominant intrinsic neuronal activations of the human brain during rest as opposed to task periods. Previous studies have demonstrated the utility of DMNs in identifying characteristic traits such as hyperactivation and hypoactivation from group-level fMRI data. However, these group-level spatial patterns (SPs) were mostly based on random-effect (RFX) statistics determined using only the intersubject variability. To reduce the potentially significant level of variability in group-level SPs in RFX due to intrasubject variability, we were motivated to adopt a mixed-effects (MFX) statistics that is using both intrasubject and intersubject variability. Publicly available group fMRI database during resting state was analyzed using a temporal concatenation-based group independent component (IC) analysis, and DMN-related ICs at the group-level were automatically selected. The individual-level SPs of these DMN-related ICs were subsequently estimated using a dual-regression approach. Using these individual-level SPs, we evaluated the reproducibility and potential variability of the DMNs from the RFX and MFX statistics using performance measures including (1) neuronal activation levels, (2) percentages of overlap, (3) Pearson's spatial correlation coefficients, and (4) the distances between center-of-clusters. The resulting SPs from the MFX-based group inference showed a significantly greater level of reproducibility than those from the RFX-based group inference as tested in a bootstrapping framework Family-wise error (FWE)-corrected p < 10−10, one-way analysis of variance (ANOVA)). The reported findings may provide a valuable supplemental option for investigating the neuropsychiatric group- or condition-dependent characteristic traits implicated in DMNs. © 2012 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 22, 121–131, 2012 © 2012 Wiley Periodicals, Inc.

For face recognition, conventional dictionary learning (DL) methods have disadvantages. In the paper, we propose a novel robust, discriminative and comprehensive DL (RDCDL) model. The proposed model uses sample diversities of the same face image to make the dictionary robust. The model includes class-specific dictionary atoms and disturbance dictionary atoms, which can well represent the data from different classes. Both the dictionary and the representation coefficients of data on the dictionary introduce discriminative information, which improves effectively the discrimination capability of the dictionary. The proposed RDCDL is extensively evaluated on benchmark face image databases, and it shows superior performance to many state-of-the-art sparse representation and dictionary learning methods for face recognition.

Autism Spectrum Disorder (ASD) is a heterogeneous disorder in children, and the current clinical diagnosis is accomplished using behavioral, cognitive, developmental, and language metrics. These clinical metrics can be imperfect measures as they are subject to high test-retest variability, and are influenced by assessment factors such as environment, social structure, or comorbid disorders. Advances in neuroimaging coupled with machine-learning provides an opportunity to develop methods that are more quantifiable, and reliable than existing clinical techniques. In this paper, we design and develop a deep-learning model that operates on functional magnetic resonance imaging (fMRI) data, and can classify between ASD and neurotypical brains. We introduce a novel strategy to transform time-series data extracted from fMRI signals into Gramian Angular Field (GAF) while locking in the temporal and spatial patterns in the data. Our motivation is to design and develop a novel framework that could encode the time-series, acquired from fMRI data, into images that can be used by deep-learning architectures that have been successful in computer vision. In our proposed framework called ASD-GResTM, we used a Convolutional Neural Network (CNN) to extract useful features from GAF images. We then used a Long Short-Term Memory (LSTM) layer to learn the activities between the regions. Finally, the output representations of the last LSTM layer are applied to a single-layer perceptron (SPL) to get the final classification. Our extensive experimentation demonstrates high accuracy across 4 centers, and outperforms state-of-the-art models on two centers with an increase in the accuracy of 17.58% and 6.7%, respectively as compared to the state of the art. Our model achieved the maximum accuracy of 81.78% with high degree of sensitivity and specificity. All training, validation, and testing was accomplished using openly available ABIDE-I benchmarking dataset.

A two-step super-Gaussian independent component analysis approach for fMRI data

Neural activity occurs in the shape of spatially organized patterns: networks of brain regions activate in synchrony. Many of these functional networks also happen to be strongly structurally connected. We use this information to revisit the fundamental problem of functional magnetic resonance imaging (fMRI) data deconvolution. Using tools from graph signal processing (GSP), we extend total activation, a spatiotemporal deconvolution technique, to data defined on graph domains. The resulting approach simultaneously cancels out the effect of the haemodynamics, and promotes spatial patterns that are in harmony with predefined structural wirings. More precisely, we minimize a functional involving one data fidelity and two regularization terms. The first regularizer uses the concept of generalized total variation to promote sparsity in the activity transients domain. The second term controls the overall spatial variation over the graph structure. We demonstrate the relevance of this structurally-driven regularization on synthetic and experimental data.

Fruit cracking and rust spots are common diseases of nectarines that seriously affect their yield and quality. Therefore, it is essential to construct fast and accurate disease-identification models for agricultural products. In this paper, a sparse dictionary learning method was proposed to realize the rapid and nondestructive identification of nectarine disease based on multiple color features combined with improved LK-SVD (Label K-Singular Value Decomposition). According to the color characteristics of the nectarine itself and the significant color differences existing in the three categories of nectarine (diseased, normal, and background parts), multiple color spaces of RGB, HSV, Lab, and YCbCr were studied. It was concluded that the G channel in RGB space, Y channel in YCbCr space, and L channel in Lab space can better distinguish the diseased part from the other parts. At the model-training stage, pixels of the diseased, normal, and background parts in the nectarine image were randomly selected as the initial training sets, and then, the neighboring image blocks of the pixels were selected to construct the feature vectors based on the above color space channels. An improved LK-SVD dictionary learning algorithm was proposed that integrated the category label into the process of dictionary learning, and thus, an over-complete feature dictionary with significant discrimination was obtained. At the model-testing stage, the orthogonal matching pursuit (OMP) algorithm was used for sparse reconstruction of the original data, which can obtain the classification categories based on the optimized feature dictionary. The experimental results show that the sparse dictionary learning method based on multi-color features combined with improved LK-SVD can identify fruit cracking and rust spot diseases of nectarines quickly and accurately, and the average overall classification accuracies were 92.06% and 88.98%, respectively, which were better than those of k-nearest neighbor (KNN), support vector machine (SVM), DeepLabV3+, and Unet++; the identification results of DeepLabV3+ and Unet++ were also relatively high, but their average time costs were much higher, requiring 126.46~265.65 s. It is demonstrated that this study can provide technical support for disease identification in agricultural products.

Many studies have shown that the independence assumption in the widely-used ICAs is not adaptive enough for brain functional networks (BFN) detection due to the complex brain hemodynamics, functional integration, artifacts embedded in functional magnetic resonance imaging (fMRI) data, etc. In this paper, inspired by sparse coding behavior of human brain, we propose an effective BFNs detection model, called sparse dictionary learning separation (SDLS). In the SDLS, facing the dilemma of huge training samples in sparse learning, an efficient spatial-domain data reduction algorithm was first designed to sharply alleviate the training cost and suppress noise. Then, an improved K singular value decomposition was proposed to speed up the correct convergence of the dictionary learning process. Furthermore, considering the variant degrees of functional integration and sparsity of BFNs across different fMRI datasets, a minimum description length-based framework was proposed to formulate two key factors, i.e., the dictionary mutual incoherence level and sparsity level, self-adaptively resulting in effective temporal dynamics model. Finally, a least-square-based functional network reconstruction was presented to extract the final BFNs. The simulated and real data experiments demonstrated that SDLS had the superiority in the spatial/temporal sources identification, and stronger spatial robustness against the variant smoothing kernels, in contrast to ICAs. SDLS was a novel data-driven BFN separation model, which had an overall consideration of multiple factors, e.g., huge samples dilemma, artifacts removal, and variant degrees of functional integration and sparsity of BFNs. SDLS as an extension to current fMRI analysis methods was a promising model, which declared the advantage of sparsity.

Task-based fMRI (tfMRI) has been widely used to study functional brain networks. Modeling tfMRI data is challenging due to at least two problems: the lack of the ground truth of underlying neural activity and the intrinsic structure of tfMRI data is highly complex. To better understand brain networks based on fMRI data, data-driven approaches were proposed, for instance, Independent Component Analysis (ICA) and Sparse Dictionary Learning (SDL). However, both ICA and SDL only build shallow models, and they are under the strong assumption that original fMRI signal could be linearly decomposed into time series components with their corresponding spatial maps. As growing evidence shows that human brain function is hierarchically organized, new approaches that can infer and model the hierarchical structure of brain networks are widely called for. Recently, deep convolutional neural network (CNN) has drawn much attention, in that deep CNN has been proven to be a powerful method for learning high-level and mid-level abstractions from low-level raw data. Inspired by the power of deep CNN, in this study, we developed a new neural network structure based on CNN, called Deep Convolutional Auto-Encoder (DCAE), in order to take the advantages of both data-driven approach and CNN’s hierarchical feature abstraction ability for the purpose of learning mid-level and high-level features from complex tfMRI time series in an unsupervised manner. The DCAE has been applied and tested on the publicly available human connectome project (HCP) tfMRI datasets, and promising results are achieved.

Multivariate analysis of fMRI time series: classification and regression of brain responses using machine learning

Slip/antiblocking agents

Cortical thickness estimation performed in vivo via magnetic resonance imaging (MRI) is an important technique for the diagnosis and understanding of the progression of Alzheimer's disease (AD). Directly using raw cortical thickness measures as features with Support Vector Machine (SVM) for clinical group classification only yields modest results since brain areas are not equally atrophied during AD progression. Therefore, feature reduction is generally required to retain only the most relevant features for the final classification. In this paper, a spherical sparse coding and dictionary learning method is proposed and it achieves relatively high classification results on publicly available data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) 2 dataset (N = 201) which contains structural MRI data of four clinical groups: cognitive unimpaired (CU), early mild cognitive impairment (EMCI), later MCI (LMCI) and AD. The proposed framework takes the estimated cortical thickness and the spherical parameterization computed by FreeSurfer as inputs and constructs weighted patches in the spherical parameter domain of the cortical surface. Then sparse coding is applied to the resulting surface patch features, followed by max-pooling to extract the final feature sets. Finally, SVM is employed for binary group classifications. The results show the superiority of the proposed method over other cortical morphometry systems and offer a different way to study the early identification and prevention of AD.