Bayesian spatiotemporal modeling on complex-valued fMRI signals via kernel convolutions.

We propose a model-based approach that combines Bayesian variable selection tools, a novel spatial kernel convolution structure, and autoregressive processes for detecting a subject's brain activation at the voxel level in complex-valued functional magnetic resonance imaging (CV-fMRI) data. A computationally efficient Markov chain Monte Carlo algorithm for posterior inference is developed by taking advantage of the dimension reduction of the kernel-based structure. The proposed spatiotemporal model leads to more accurate posterior probability activation maps and less false positives than alternative spatial approaches based on Gaussian process models, and other complex-valued models that do not incorporate spatial and/or temporal structure. This is illustrated in the analysis of simulated data and human task-related CV-fMRI data. In addition, we show that complex-valued approaches dominate magnitude-only approaches and that the kernel structure in our proposed model considerably improves sensitivity rates when detecting activation at the voxel level.

Topological Data Analysis for Multivariate Time Series Data.

Over the last two decades, topological data analysis (TDA) has emerged as a very powerful data analytic approach that can deal with various data modalities of varying complexities. One of the most commonly used tools in TDA is persistent homology (PH), which can extract topological properties from data at various scales. The aim of this article is to introduce TDA concepts to a statistical audience and provide an approach to analyzing multivariate time series data. The application's focus will be on multivariate brain signals and brain connectivity networks. Finally, this paper concludes with an overview of some open problems and potential application of TDA to modeling directionality in a brain network, as well as the casting of TDA in the context of mixed effect models to capture variations in the topological properties of data collected from multiple subjects.

Estimation and probabilistic projection of levels and trends in the sex ratio at birth in seven provinces of Nepal from 1980 to 2050: a Bayesian modeling approach.

BACKGROUND: The sex ratio at birth (SRB; ratio of male to female births) in Nepal has been reported around the normal level on the national level. However, the national SRB could mask the disparity within the country. Given the demographic and cultural heterogeneities in Nepal, it is crucial to model Nepal SRB on the subnational level. Prior studies on subnational SRB in Nepal are mostly based on reporting observed values from surveys and census, and no study has provided probabilistic projections. We aim to estimate and project SRB for the seven provinces of Nepal from 1980 to 2050 using a Bayesian modeling approach. METHODS: We compiled an extensive database on provincial SRB of Nepal, consisting 2001, 2006, 2011, and 2016 Nepal Demographic and Health Surveys and 2011 Census. We adopted a Bayesian hierarchical time series model to estimate and project the provincial SRB, with a focus on modelling the potential SRB imbalance. RESULTS: In 2016, the highest SRB is estimated in Province 5 (Lumbini Pradesh) at 1.102, corresponding to 110.2 male births per 100 female births, with a 95% credible interval (1.044, 1.127) and the lowest SRB is in Province 2 at 1.053 (1.035, 1.109). The SRB imbalance probabilities in all provinces are generally low and vary from 16% in Province 2 to 81% in Province 5 (Lumbini Pradesh). SRB imbalances are estimated to have begun at the earliest in 2001 in Province 5 (Lumbini Pradesh) with a 95% credible interval (1992, 2022) and the latest in 2017 (1998, 2040) in Province 2. We project SRB in all provinces to begin converging back to the national baseline in the mid-2030s. By 2050, the SRBs in all provinces are projected to be around the SRB baseline level. CONCLUSIONS: Our findings imply that the majority of provinces in Nepal have low risks of SRB imbalance for the period 1980-2016. However, we identify a few provinces with higher probabilities of having SRB inflation. The projected SRB is an important illustration of potential future prenatal sex discrimination and shows the need to monitor SRB in provinces with higher possibilities of SRB imbalance.

Brain Waves Analysis Via a Non-Parametric Bayesian Mixture of Autoregressive Kernels.

The standard approach to analyzing brain electrical activity is to examine the spectral density function (SDF) and identify frequency bands, defined a priori, that have the most substantial relative contributions to the overall variance of the signal. However, a limitation of this approach is that the precise frequency and bandwidth of oscillations are not uniform across different cognitive demands. Thus, these bands should not be arbitrarily set in any analysis. To overcome this limitation, the Bayesian mixture auto-regressive decomposition (BMARD) method is proposed, as a data-driven approach that identifies (i) the number of prominent spectral peaks, (ii) the frequency peak locations, and (iii) their corresponding bandwidths (or spread of power around the peaks). Using the BMARD method, the standardized SDF is represented as a Dirichlet process mixture based on a kernel derived from second-order auto-regressive processes which completely characterize the location (peak) and scale (bandwidth) parameters. A Metropolis-Hastings within the Gibbs algorithm is developed for sampling the posterior distribution of the mixture parameters. Simulations demonstrate the robust performance of the proposed method. Finally, the BMARD method is applied to analyze local field potential (LFP) activity from the hippocampus of laboratory rats across different conditions in a non-spatial sequence memory experiment, to identify the most prominent frequency bands and examine the link between specific patterns of brain oscillatory activity and trial-specific cognitive demands.

Regularized matrix data clustering and its application to image analysis.

We propose a novel regularized mixture model for clustering matrix-valued data. The proposed method assumes a separable covariance structure for each cluster and imposes a sparsity structure (eg, low rankness, spatial sparsity) for the mean signal of each cluster. We formulate the problem as a finite mixture model of matrix-normal distributions with regularization terms, and then develop an expectation maximization type of algorithm for efficient computation. In theory, we show that the proposed estimators are strongly consistent for various choices of penalty functions. Simulation and two applications on brain signal studies confirm the excellent performance of the proposed method including a better prediction accuracy than the competitors and the scientific interpretability of the solution.

Ridge-penalized adaptive Mantel test and its application in imaging genetics.

We propose a ridge-penalized adaptive Mantel test (AdaMant) for evaluating the association of two high-dimensional sets of features. By introducing a ridge penalty, AdaMant tests the association across many metrics simultaneously. We demonstrate how ridge penalization bridges Euclidean and Mahalanobis distances and their corresponding linear models from the perspective of association measurement and testing. This result is not only theoretically interesting but also has important implications in penalized hypothesis testing, especially in high-dimensional settings such as imaging genetics. Applying the proposed method to an imaging genetic study of visual working memory in healthy adults, we identified interesting associations of brain connectivity (measured by electroencephalogram coherence) with selected genetic features.

RITS: a toolbox for assessing complex interventions via interrupted time series models.

BACKGROUND: Various interacting and interdependent components comprise complex interventions. These components create difficulty in assessing the true impact of interventions designed to improve patient-centered outcomes. Interrupted time series (ITS) designs borrow from case-crossover designs and serve as quasi-experimental methodology able to retrospectively assess the impact of an intervention while accounting for temporal correlation. While ITS designs are aptly situated for studying the impacts of large-scale public health policies, existing ITS software implement rigid ITS methodology that often assume the pre- and post-intervention phases are fully differentiated (by a known change-point or set of time points) and do not allow for changes in both the mean functions and correlation structure. RESULTS: This article describes the Robust Interrupted Time Series (RITS) toolbox, a stand-alone user-friendly application researchers can use to implement flexible ITS models that estimate the lagged effect of an intervention on an outcome, level and trend changes, and post-intervention changes in the correlation structure, for single and multiple ITS. The RITS toolbox incorporates a formal test for the existence of a change in the outcome and estimates a change-point over a set of possible change-points defined by the researcher. In settings with multiple ITS, RITS provides a global over-all units change-point and allows for unit-specific changes in the mean functions and correlation structures. CONCLUSIONS: The RITS toolbox is the first piece of software that allows researchers to use flexible ITS models that test for the existence of a change-point, estimate the change-point (if estimation is desired), and allow for changes in both the mean functions and correlation structures at the change point. RITS does not require any knowledge of a statistical (or otherwise) programming language, is freely available to the community, and may be downloaded and used on a local machine to ensure data protection.

Multi-Chaotic Analysis of Inter-Beat (R-R) Intervals in Cardiac Signals for Discrimination between Normal and Pathological Classes.

Cardiac signals have complex structures representing a combination of simpler structures. In this paper, we develop a new data analytic tool that can extract the complex structures of cardiac signals using the framework of multi-chaotic analysis, which is based on the p-norm for calculating the largest Lyapunov exponent (LLE). Appling the p-norm is useful for deriving the spectrum of the generalized largest Lyapunov exponents (GLLE), which is characterized by the width of the spectrum (which we denote by W). This quantity measures the degree of multi-chaos of the process and can potentially be used to discriminate between different classes of cardiac signals. We propose the joint use of the GLLE and spectrum width to investigate the multi-chaotic behavior of inter-beat (R-R) intervals of cardiac signals recorded from 54 healthy subjects (hs), 44 subjects diagnosed with congestive heart failure (chf), and 25 subjects diagnosed with atrial fibrillation (af). With the proposed approach, we build a regression model for the diagnosis of pathology. Multi-chaotic analysis showed a good performance, allowing the underlying dynamics of the system that generates the heart beat to be examined and expert systems to be built for the diagnosis of cardiac pathologies.

Evolutionary State-Space Model and Its Application to Time-Frequency Analysis of Local Field Potentials.

We propose an evolutionary state space model (E-SSM) for analyzing high dimensional brain signals whose statistical properties evolve over the course of a non-spatial memory experiment. Under E-SSM, brain signals are modeled as mixtures of components (e.g., AR(2) process) with oscillatory activity at pre-defined frequency bands. To account for the potential non-stationarity of these components (since the brain responses could vary throughout the entire experiment), the parameters are allowed to vary over epochs. Compared with classical approaches such as independent component analysis and filtering, the proposed method accounts for the entire temporal correlation of the components and accommodates non-stationarity. For inference purpose, we propose a novel computational algorithm based upon using Kalman smoother, maximum likelihood and blocked resampling. The E-SSM model is applied to simulation studies and an application to a multi-epoch local field potentials (LFP) signal data collected from a non-spatial (olfactory) sequence memory task study. The results confirm that our method captures the evolution of the power for different components across different phases in the experiment and identifies clusters of electrodes that behave similarly with respect to the decomposition of different sources. These findings suggest that the activity of different electrodes does change over the course of an experiment in practice; treating these epoch recordings as realizations of an identical process could lead to misleading results. In summary, the proposed method underscores the importance of capturing the evolution in brain responses over the study period.

Modeling Spectral Properties in Stationary Processes of Varying Dimensions with Applications to Brain Local Field Potential Signals.

In some applications, it is important to compare the stochastic properties of two multivariate time series that have unequal dimensions. A new method is proposed to compare the spread of spectral information in two multivariate stationary processes with different dimensions. To measure discrepancies, a frequency specific spectral ratio (FS-ratio) statistic is proposed and its asymptotic properties are derived. The FS-ratio is blind to the dimension of the stationary process and captures the proportion of spectral power in various frequency bands. Here we develop a technique to automatically identify frequency bands that carry significant spectral power. We apply our method to track changes in the complexity of a 32-channel local field potential (LFP) signal from a rat following an experimentally induced stroke. At every epoch (a distinct time segment from the duration of the experiment), the nonstationary LFP signal is decomposed into stationary and nonstationary latent sources and the complexity is analyzed through these latent stationary sources and their dimensions that can change across epochs. The analysis indicates that spectral information in the Beta frequency band (12-30 Hertz) demonstrated the greatest change in structure and complexity due to the stroke.

Assessing health care interventions via an interrupted time series model: Study power and design considerations.

The delivery and assessment of quality health care is complex with many interacting and interdependent components. In terms of research design and statistical analysis, this complexity and interdependency makes it difficult to assess the true impact of interventions designed to improve patient health care outcomes. Interrupted time series (ITS) is a quasi-experimental design developed for inferring the effectiveness of a health policy intervention while accounting for temporal dependence within a single system or unit. Current standardized ITS methods do not simultaneously analyze data for several units nor are there methods to test for the existence of a change point and to assess statistical power for study planning purposes in this context. To address this limitation, we propose the "Robust Multiple ITS" (R-MITS) model, appropriate for multiunit ITS data, that allows for inference regarding the estimation of a global change point across units in the presence of a potentially lagged (or anticipatory) treatment effect. Under the R-MITS model, one can formally test for the existence of a change point and estimate the time delay between the formal intervention implementation and the over-all-unit intervention effect. We conducted empirical simulation studies to assess the type one error rate of the testing procedure, power for detecting specified change-point alternatives, and accuracy of the proposed estimating methodology. R-MITS is illustrated by analyzing patient satisfaction data from a hospital that implemented and evaluated a new care delivery model in multiple units.

System- and Unit-Level Care Quality Outcome Improvements After Integrating Clinical Nurse Leaders Into Frontline Care Delivery.

OBJECTIVE: This study determined whether 1 health system's frontline nursing model redesign to integrate clinical nurse leaders (CNLs) improved care quality and outcome score consistency. METHODS: Interrupted time-series design was used to measure patient satisfaction with 7 metrics before and after formally integrating CNLs into a Michigan healthcare system. Analysis generated estimates of quality outcome: a) change point; b) level change; and c) variance, pre-post implementation. RESULTS: The lowest-performing unit showed significant increases in quality scores, but there were no significant increases at the hospital level. Quality metric consistency increased significantly for every indicator at the hospital and unit level. CONCLUSIONS: To our knowledge, this is the 1st study quantifying quality outcome consistency before and after nursing care delivery redesign with CNLs. The significant improvement suggests the CNL care model is associated with production of stable clinical microsystem practices that help to reduce clinical variability, thus improving care quality.

Statistical models for brain signals with properties that evolve across trials.

Most neuroscience cognitive experiments involve repeated presentations of various stimuli across several minutes or a few hours. It has been observed that brain responses, even to the same stimulus, evolve over the course of the experiment. These changes in brain activation and connectivity are believed to be associated with learning and/or habituation. In this paper, we present two general approaches to modeling dynamic brain connectivity using electroencephalograms (EEGs) recorded across replicated trials in an experiment. The first approach is the Markovian regime-switching vector autoregressive model (MS-VAR) which treats EEGs as realizations of an underlying brain process that switches between different states both within a trial and across trials in the entire experiment. The second is the slowly evolutionary locally stationary process (SEv-LSP) which characterizes the observed EEGs as a mixture of oscillatory activities at various frequency bands. The SEv-LSP model captures the dynamic nature of the amplitudes of the band-oscillations and cross-correlations between them. The MS-VAR model is able to capture abrupt changes in the dynamics while the SEv-LSP directly gives interpretable results. Moreover, it is nonparametric and hence does not suffer from model misspecification. For both of these models, time-evolving connectivity metrics in the frequency domain are derived from the model parameters for both functional and effective connectivity. We illustrate these two models for estimating cross-trial connectivity in selective attention using EEG data from an oddball paradigm auditory experiment where the goal is to characterize the evolution of brain responses to target stimuli and to standard tones presented randomly throughout the entire experiment. The results suggest dynamic changes in connectivity patterns over trials with inter-subject variability.

A scalable multi-resolution spatio-temporal model for brain activation and connectivity in fMRI data.

Functional Magnetic Resonance Imaging (fMRI) is a primary modality for studying brain activity. Modeling spatial dependence of imaging data at different spatial scales is one of the main challenges of contemporary neuroimaging, and it could allow for accurate testing for significance in neural activity. The high dimensionality of this type of data (on the order of hundreds of thousands of voxels) poses serious modeling challenges and considerable computational constraints. For the sake of feasibility, standard models typically reduce dimensionality by modeling covariance among regions of interest (ROIs)-coarser or larger spatial units-rather than among voxels. However, ignoring spatial dependence at different scales could drastically reduce our ability to detect activation patterns in the brain and hence produce misleading results. We introduce a multi-resolution spatio-temporal model and a computationally efficient methodology to estimate cognitive control related activation and whole-brain connectivity. The proposed model allows for testing voxel-specific activation while accounting for non-stationary local spatial dependence within anatomically defined ROIs, as well as regional dependence (between-ROIs). The model is used in a motor-task fMRI study to investigate brain activation and connectivity patterns aimed at identifying associations between these patterns and regaining motor functionality following a stroke.

Spectral synchronicity in brain signals.

This paper addresses the problem of identifying brain regions with similar oscillatory patterns detected from electroencephalograms. We introduce the hierarchical spectral merger (HSM) clustering method where the feature of interest is the spectral curve and the similarity metric used is the total variance distance. The HSM method is compared with clustering using features derived from independent-component analysis. Moreover, the HSM method is applied to 2 different electroencephalogram datasets. The first was recorded at resting state where the participant was not engaged in any cognitive task; the second was recorded during a spontaneous epileptic seizure. The results of the analyses using the HSM method demonstrate that clustering could evolve over the duration of the resting state and during epileptic seizure.

A robust interrupted time series model for analyzing complex health care intervention data.

Current health policy calls for greater use of evidence-based care delivery services to improve patient quality and safety outcomes. Care delivery is complex, with interacting and interdependent components that challenge traditional statistical analytic techniques, in particular, when modeling a time series of outcomes data that might be "interrupted" by a change in a particular method of health care delivery. Interrupted time series (ITS) is a robust quasi-experimental design with the ability to infer the effectiveness of an intervention that accounts for data dependency. Current standardized methods for analyzing ITS data do not model changes in variation and correlation following the intervention. This is a key limitation since it is plausible for data variability and dependency to change because of the intervention. Moreover, present methodology either assumes a prespecified interruption time point with an instantaneous effect or removes data for which the effect of intervention is not fully realized. In this paper, we describe and develop a novel robust interrupted time series (robust-ITS) model that overcomes these omissions and limitations. The robust-ITS model formally performs inference on (1) identifying the change point; (2) differences in preintervention and postintervention correlation; (3) differences in the outcome variance preintervention and postintervention; and (4) differences in the mean preintervention and postintervention. We illustrate the proposed method by analyzing patient satisfaction data from a hospital that implemented and evaluated a new nursing care delivery model as the intervention of interest. The robust-ITS model is implemented in an R Shiny toolbox, which is freely available to the community.

Simultaneous control of error rates in fMRI data analysis.

The key idea of statistical hypothesis testing is to fix, and thereby control, the Type I error (false positive) rate across samples of any size. Multiple comparisons inflate the global (family-wise) Type I error rate and the traditional solution to maintaining control of the error rate is to increase the local (comparison-wise) Type II error (false negative) rates. However, in the analysis of human brain imaging data, the number of comparisons is so large that this solution breaks down: the local Type II error rate ends up being so large that scientifically meaningful analysis is precluded. Here we propose a novel solution to this problem: allow the Type I error rate to converge to zero along with the Type II error rate. It works because when the Type I error rate per comparison is very small, the accumulation (or global) Type I error rate is also small. This solution is achieved by employing the likelihood paradigm, which uses likelihood ratios to measure the strength of evidence on a voxel-by-voxel basis. In this paper, we provide theoretical and empirical justification for a likelihood approach to the analysis of human brain imaging data. In addition, we present extensive simulations that show the likelihood approach is viable, leading to "cleaner"-looking brain maps and operational superiority (lower average error rate). Finally, we include a case study on cognitive control related activation in the prefrontal cortex of the human brain.

A semiparametric Bayesian model for detecting synchrony among multiple neurons.

We propose a scalable semiparametric Bayesian model to capture dependencies among multiple neurons by detecting their cofiring (possibly with some lag time) patterns over time. After discretizing time so there is at most one spike at each interval, the resulting sequence of 1s (spike) and 0s (silence) for each neuron is modeled using the logistic function of a continuous latent variable with a gaussian process prior. For multiple neurons, the corresponding marginal distributions are coupled to their joint probability distribution using a parametric copula model. The advantages of our approach are as follows. The nonparametric component (i.e., the gaussian process model) provides a flexible framework for modeling the underlying firing rates, and the parametric component (i.e., the copula model) allows us to make inferences regarding both contemporaneous and lagged relationships among neurons. Using the copula model, we construct multivariate probabilistic models by separating the modeling of univariate marginal distributions from the modeling of a dependence structure among variables. Our method is easy to implement using a computationally efficient sampling algorithm that can be easily extended to high-dimensional problems. Using simulated data, we show that our approach could correctly capture temporal dependencies in firing rates and identify synchronous neurons. We also apply our model to spike train data obtained from prefrontal cortical areas.

Unleashing the potential of fNIRS with machine learning: classification of fine anatomical movements to empower future brain-computer interface.

In this study, we explore the potential of using functional near-infrared spectroscopy (fNIRS) signals in conjunction with modern machine-learning techniques to classify specific anatomical movements to increase the number of control commands for a possible fNIRS-based brain-computer interface (BCI) applications. The study focuses on novel individual finger-tapping, a well-known task in fNIRS and fMRI studies, but limited to left/right or few fingers. Twenty-four right-handed participants performed the individual finger-tapping task. Data were recorded by using sixteen sources and detectors placed over the motor cortex according to the 10-10 international system. The event's average oxygenated Δ HbO and deoxygenated Δ HbR hemoglobin data were utilized as features to assess the performance of diverse machine learning (ML) models in a challenging multi-class classification setting. These methods include LDA, QDA, MNLR, XGBoost, and RF. A new DL-based model named "Hemo-Net" has been proposed which consists of multiple parallel convolution layers with different filters to extract the features. This paper aims to explore the efficacy of using fNRIS along with ML/DL methods in a multi-class classification task. Complex models like RF, XGBoost, and Hemo-Net produce relatively higher test set accuracy when compared to LDA, MNLR, and QDA. Hemo-Net has depicted a superior performance achieving the highest test set accuracy of 76%, however, in this work, we do not aim at improving the accuracies of models rather we are interested in exploring if fNIRS has the neural signatures to help modern ML/DL methods in multi-class classification which can lead to applications like brain-computer interfaces. Multi-class classification of fine anatomical movements, such as individual finger movements, is difficult to classify with fNIRS data. Traditional ML models like MNLR and LDA show inferior performance compared to the ensemble-based methods of RF and XGBoost. DL-based method Hemo-Net outperforms all methods evaluated in this study and demonstrates a promising future for fNIRS-based BCI applications.

Graph Autoencoders for Embedding Learning in Brain Networks and Major Depressive Disorder Identification.

Brain functional connectivity (FC) networks inferred from functional magnetic resonance imaging (fMRI) have shown altered or aberrant brain functional connectome in various neuropsychiatric disorders. Recent application of deep neural networks to connectome-based classification mostly relies on traditional convolutional neural networks (CNNs) using input FCs on a regular Euclidean grid to learn spatial maps of brain networks neglecting the topological information of the brain networks, leading to potentially sub-optimal performance in brain disorder identification. We propose a novel graph deep learning framework that leverages non-Euclidean information inherent in the graph structure for classifying brain networks in major depressive disorder (MDD). We introduce a novel graph autoencoder (GAE) architecture, built upon graph convolutional networks (GCNs), to embed the topological structure and node content of large fMRI networks into low-dimensional representations. For constructing the brain networks, we employ the Ledoit-Wolf (LDW) shrinkage method to efficiently estimate high-dimensional FC metrics from fMRI data. We explore both supervised and unsupervised techniques for graph embedding learning. The resulting embeddings serve as feature inputs for a deep fully-connected neural network (FCNN) to distinguish MDD from healthy controls (HCs). Evaluating our model on resting-state fMRI MDD dataset, we observe that the GAE-FCNN outperforms several state-of-the-art methods for brain connectome classification, achieving the highest accuracy when using LDW-FC edges as node features. The graph embeddings of fMRI FC networks also reveal significant group differences between MDD and HCs. Our framework demonstrates the feasibility of learning graph embeddings from brain networks, providing valuable discriminative information for diagnosing brain disorders.