Bayesian Image Mediation Analysis

Yuliang Xu,Jian Kang
2023-10-25
Abstract:Mediation analysis aims to separate the indirect effect through mediators from the direct effect of the exposure on the outcome. It is challenging to perform mediation analysis with neuroimaging data which involves high dimensionality, complex spatial correlations, sparse activation patterns and relatively low signal-to-noise ratio. To address these issues, we develop a new spatially varying coefficient structural equation model for Bayesian Image Mediation Analysis (BIMA). We define spatially varying mediation effects within the potential outcome framework, employing the soft-thresholded Gaussian process prior for functional parameters. We establish the posterior consistency for spatially varying mediation effects along with selection consistency on important regions that contribute to the mediation effects. We develop an efficient posterior computation algorithm scalable to analysis of large-scale imaging data. Through extensive simulations, we show that BIMA can improve the estimation accuracy and computational efficiency for high-dimensional mediation analysis over the existing methods. We apply BIMA to analyze the behavioral and fMRI data in the Adolescent Brain Cognitive Development (ABCD) study with a focus on inferring the mediation effects of the parental education level on the children's general cognitive ability that are mediated through the working memory brain activities.
Methodology,Statistics Theory,Computation
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the challenge of conducting mediation analysis in high - dimensional neuroimaging data. Specifically, the authors developed a new Spatially Varying Coefficient Structural Equation Model (SV - CSEM) for Bayesian Image Mediation Analysis (BIMA). This method aims to deal with the high - dimensionality, complex spatial correlations, sparse activation patterns, and low signal - to - noise ratio in neuroimaging data. ### Background and Problem of the Paper In neuroimaging research, Mediation Analysis is an important statistical tool that decomposes the total effect of an exposure or treatment variable on an outcome variable into a direct effect and an indirect effect through a mediator variable. However, using neuroimaging data for mediation analysis faces several challenges: 1. **High - dimensionality**: The number of voxels in a standard brain template can reach 200,000, which requires a large amount of computational resources to implement statistical algorithms. 2. **Complex spatial correlations**: There are complex correlation patterns among brain - imaging mediator variables, such as correlations between neighboring voxels and correlations between brain regions with the same function. Ignoring or inappropriately handling these correlations may introduce bias or lose statistical efficiency. 3. **Low signal - to - noise ratio**: Due to the low signal - to - noise ratio of brain - imaging data, voxel - level imaging mediator variables may have a weak or zero effect on the outcome variable, resulting in low power and a high false - positive rate when detecting active mediator variables. ### Solution To solve the above problems, the authors proposed the BIMA method, whose main features include: - **Spatially varying coefficient model**: Defined the spatially varying mediation effect in the potential outcome framework and adopted the Soft - thresholded Gaussian Process (STGP) to handle functional parameters. - **Posterior consistency**: Established the posterior consistency of the spatially varying mediation effect and also has consistency in the selection of important regions. - **Efficient posterior computation algorithm**: Developed an efficient posterior computation algorithm that can be scaled to large - scale imaging data analysis. ### Application Example The authors applied BIMA to the behavioral and fMRI data in the Adolescent Brain Cognitive Development (ABCD) study, focusing on how parental education level affects children's general cognitive ability through working - memory brain activity. ### Main Contributions - **Theoretical contribution**: Established the posterior consistency of the spatially varying coefficient, ensuring the accuracy of the estimation of the spatially varying mediation effect. - **Methodological innovation**: Proposed a new Spatially Varying Coefficient Structural Equation Model that can handle high - dimensional and complex - correlated neuroimaging data. - **Practical application**: Verified the effectiveness and computational efficiency of the BIMA method in real - data, especially outperforming existing methods in high - dimensional mediation analysis. ### Conclusion The BIMA method provides a powerful tool for mediation analysis of high - dimensional neuroimaging data, which can effectively deal with problems such as high - dimensionality, complex correlations, and low signal - to - noise ratio, providing new perspectives and methods for neuroimaging research.