• Corpus ID: 247011227

Image Response Regression via Deep Neural Networks

  title={Image Response Regression via Deep Neural Networks},
  author={Daiwei Zhang and Lexin Li and Chandra Sekhar Sripada and Jian Kang},
Delineating the associations between images and a vector of covariates is of central interest in medical imaging studies. To tackle this problem of image response regression, we propose a novel nonparametric approach in the framework of spatially varying coefficient models, where the spatially varying functions are estimated through deep neural networks. Compared to existing solutions, the proposed method explicitly accounts for spatial smoothness and subject heterogeneity, has straightforward… 

Figures and Tables from this paper

Density Regression and Uncertainty Quantification with Bayesian Deep Noise Neural Networks

This work proposes the Bayesian Deep Noise Neural Network (B-DeepNoise), which generalizes standard Bayesian DNNs by extending the random noise variable from the output layer to all hidden layers, and applies it to predict general intelligence from neuroimaging features in the Adolescent Brain Cognitive Development project.



Sparse Learning and Structure Identification for Ultrahigh-Dimensional Image-on-Scalar Regression

This article develops a unified approach for simultaneous sparse learning and model structure identification in the presence of ultrahigh-dimensional covariates via a flexible partially linear spatially varying coefficient model for high-dimensional image-on-scalar regression.

Multivariate Spline Estimation and Inference for Image-on-Scalar Regression

: Motivated by recent analyses of data in biomedical imaging studies, we consider a class of image-on-scalar regression models for imaging responses and scalar predictors. We propose using flexible

Local Region Sparse Learning for Image-on-Scalar Regression

A novel region-selection penalty is introduced in the framework of image-on-scalar regression that combines the Smoothly Clipped Absolute Deviation regularization, enforcing sparsity, and the SCAD of total variation (TV) regularization into one group, which segments contiguous spatial regions against zero-valued background.

Spatially Varying Coefficient Model for Neuroimaging Data With Jump Discontinuities

A novel spatially varying coefficient model (SVCM) is proposed to capture the varying association between imaging measures in a three-dimensional volume (or two-dimensional surface) with a set of covariates to investigate the asymptotic properties of the multiscale adaptive parameter estimates.

Thresholded Multiscale Gaussian Processes with Application to Bayesian Feature Selection for Massive Neuroimaging Data

A spatially varying coefficient model (SVCMs) with sparsity and piecewise smoothness imposed on the coefficient functions is proposed and a new class of nonparametric priors based on thresholded multiresolution Gaussian processes (TMGP) is developed.

STORE: Sparse Tensor Response Regression and Neuroimaging Analysis

A new regression model with a tensor response and a vector predictor, Sparse TensOr REsponse regression (STORE), with a non-asymptotic estimation error bound is proposed, and an interesting interaction between the computational efficiency and the statistical rate of convergence is revealed.

Parsimonious Tensor Response Regression

P parsimonious tensor response regression adopting a generalized sparsity principle is proposed, which effectively reduces the number of free parameters, leading to feasible computation and improved interpretation, and it enjoys a competitive finite sample performance.

A selective overview of feature screening methods with applications to neuroimaging data

A selective review of the recent developments in ultrahigh-dimensional variable screening, with a focus on their practical performance on the analysis of neuroimaging data with complex spatial correlation structures and high-dimensionality.

Adaptive spatial smoothing of fMRI images

This paper introduces a novel approach towards spatially smooth fMRI data based on the use of nonstationaryspatial Gaussian Markov random randomness (Yue and Speckman,2009), which allows the amount of smooth-ing to vary across the brain depending on the spatial ex-tent of activation, but also enables researchers to study how the extent of activation changes over time.

A Functional Varying-Coefficient Single-Index Model for Functional Response Data

A novel functional varying-coefficient single-index model (FVCSIM) is proposed to carry out the regression analysis of functional response data on a set of covariates of interest to investigate the development of white matter diffusivities along the corpus callosum skeleton obtained from Alzheimer’s Disease Neuroimaging Initiative study.