Joint Analysis of Neuroimaging Data: High-Dimensional Problems, Spatiotemporal Models and Computation

Why Study Neuroimaging Data?

Uncovering the mysteries of the brain, including its function and structure, is a key challenge of modern science. With recent advances in the speed and accuracy of data acquisition from neuroimaging tools*, we can gain a better understanding of brain health and disease.

In early neuroimaging, most studies were focused on detecting activated regions of the brain using a single neuroimaging modality. Recently, researchers have turned their attention to more complicated problems that involve integrating complementary sources of information.

The development of statistical methods for these problems has fallen behind the technological advances that allow us to collect the data. This project will bring together researchers in various disciplines to develop, test, apply, and propagate new methods. 

*tools such as functional Magnetic Resonance Imaging (fMRI), Diffusion Tensor Imaging (DTI), Electroencephalography (EEG), and Magnetoencephalography (MEG).

Areas of Exploration

Regression Modeling

Includes developing a sparse projection regression modeling framework for the high-dimensional analysis of combined neuroimaging and genomic (SNP) data.

Solutions for NIP

Includes developing computationally-feasible solutions for the neuroelectromagnetic inverse problem (NIP). This is based on combined magnetoencephalography (MEG) and electroencephalography (​​EEG) data.

The Gut-Brain Connection

Includes investigating the physiological connection between the gastrointestinal tract and the brain. This is achieved using metagenomics data to model the relationship between neural outcomes and the human intestinal microbiome.

Solving Global Challenges

Research Team’s Goal

To develop solutions for analyzing and modeling brain imaging and genomic data from multiple sources. This data will be used to investigate physiological connections between the brain and intestinal microbiome.