Section on Statistics in Imaging Student Paper Competition Award Winners

Structural brain changes (such as alterations in volume and area) are typically associated with normal brain aging. Therefore, to monitor brain health, it is crucial to monitor the geometric variations of subcortical brain structures as early as possible. However, current statistical approaches in modeling such variations face several challenges, including (i) the non-Euclidean representation of 3D shapes; (ii) the complex spatial correlation structure in local geometry; (iii) subject-level imaging heterogeneity due to misalignment of shapes in imaging preprocessing steps; (iv) group-level imaging heterogeneity due to distinct brain aging patterns within normal controls; and (v) geometric variations associated with covariates of interest (e.g., gender and education length), which may be high-dimensional.
To address these challenges, we propose a Clusterwise Shape-on-scalar FActor Regression Model (CS-FARM). In each cluster, a geodesic regression structure including covariates of interest and alignment step is established along with the Riemannian Gaussian distribution in the pre-shape space, and a latent factor model is built in the tangent space. A penalized likelihood approach is used to implement the variable selection in CS-FARM. In addition, a Monte Carlo EM algorithm is provided for the parameter estimation procedure. Finally, both simulation studies and real data analysis based on 3D brain subcortical structures from two brain aging imaging studies are conducted to evaluate the finite sample performance of CS-FARM. 

Antidepressant treatment response remains highly unpredictable, with moderate
remission rates due to heterogeneous disease mechanisms. Electroencephalography
(EEG) biomarkers has shown utilities for guiding optimal treatment rule but poses
challenges for direct use due to the need for manual feature extraction and lack of
interpretability for its complex high-dimensional property. In this paper, we propose a
deep representation learning framework EEG-Variational Autoencoder (EEGVAE) ar-
chitecture that incorporates a convolution neural network based encoder base designed
for EEG (i.e., EEGNet), and extends to a multi-head EEGVAE variant for estimating
heterogeneous treatment effects (HTE) to improve treatment response prediction. Our
approach offers several advantages by providing 1) an end-to-end learning framework
to eliminate the need for manual feature engineering; 2) flexibility in incorporating
multi-modality data; 3) the ability to learn robust, interpretable EEG representation.
Through simulations based on real data and application to a large randomized clin-
ical trial EMBARC, we demonstrate our method's superior performance in predict-
ing responder status and optimal treatment policy estimation compared to existing
approaches. The extracted latent encoding provides insights into spatial and tempo-
ral EEG patterns that influence treatment outcomes, advancing our understanding of
treatment response mechanisms in depression. 

Understanding the interplay between high-dimensional data from different views is essential in biomedical research, in fields like genomics and neuroimaging. Existing statistical association tests for two random vectors often do not fully capture dependencies between views due to limitations in modeling within-view dependencies, particularly in unstructured high-dimensional data without clear dependency patterns, leading to a potential loss of statistical power. In this work, we propose a novel approach termed devariation which is considered as a simple yet effective preprocessing method to address the limitations by adopting a penalized low-rank factor model
to flexibly capture within-view dependencies. Theoretical asymptotic power analysis shows that devariation increases statistical power, especially when within-view correlations impact signal-to-noise ratios, while maintaining robustness in Scenarios without strong internal correlations. Simulation studies highlight devariation's superior performance over existing methods in various Scenarios. We further validated devariation in neuroimaging data from the UK Biobank study, examining the associations between imaging-driven phenotypes (IDPs) derived from functional, structural, and diffusion magnetic resonance imaging (MRI). 

Magnetic resonance imaging has significantly improved our understanding of the connectivity patterns within the human brain by enabling measurement of the strength of anatomical connections between brain regions through white matter fibers (structural connectivity) and the degree of coactivation of brain regions (functional connectivity). Heritability analyses of connectivity have established that genetics account for a considerable portion of the observed intersubject variability. However, such analyses typically ignore the multidimensional nature of functional and structural connectomes. In this work, we model observed brain connectivity as the sum of multidimensional latent genetic and environmental contributions and introduce a novel constrained estimator for the covariance matrices of the genetic and environmental components. Our estimator is several orders of magnitude faster than existing methods without sacrificing estimation accuracy. The proposed covariance estimate provides a summary statistic which can be used to estimate the parameters of a novel regression analysis that enables us to characterize the relationship between the latent genetic components of structural and functional connectomes. Our analysis suggests that the genetic component of functional connectomes is highly predictable from the genetic component of structural connectomes, suggesting a close relationship at the genetic level that is attenuated by distinct environmental factors. 

The goal in this paper is to develop a novel statistical approach to characterize functional interactions between channels in a brain network. Wavelets are effective for capturing transient properties of nonstationary signals because they have compact support that can be compressed or stretched according to the dynamic properties of the signal. Wavelets give a multi-scale decomposition of signals and thus can be used for studying potential cross-scale interactions between signals. To achieve this, we develop the scale-specific subprocesses of a multivariate locally stationary wavelet stochastic process. Under this proposed framework, a novel cross-scale dependence measure is developed. This provides a measure for the dependence structure of components at different scales of multivariate time series. Extensive simulation studies are conducted to demonstrate that the theoretical properties hold in practice. The proposed cross-scale analysis is applied to the electroencephalogram (EEG) data to study alterations in the functional connectivity structure in children diagnosed with attention deficit hyperactivity disorder (ADHD). Our approach identified novel interesting cross-scale interactions between channels in the brain network.