Analyzing Neuroimaging Data

Farouk Nathoo, a University of Victoria statistician and Canada Research Chair in Biostatistics for Spatial and High-Dimensional Data, is working with his Collaborative Research Team (CRT) in developing a methodology to determine how genetic markers influence the structure and function of the brain. Co-led by University of Alberta statistician Linglong Kong, their CRT is entitled Joint Analysis of Neuroimaging Data: High-Dimensional Problems, Spatiotemporal Models and Computation. This team’s innovative neuroimaging work helps scientists better understand the function of the brain after diseases and improves their understanding of how the brain learns.

Artificially Coloured MRI Scan Of Human Brain

As neuroimaging studies collect an increasingly large volume of complex data, increased interdisciplinary collaboration between neuroscientists and statisticians is critical. Neuroscience is increasingly relying on new statistical tools and methods to unlock new insights in their data. The Canadian Statistical Sciences Institute (CANSSI) is supporting this CRT and similar collaborations across Canada in an effort to emphasize the co-creation of knowledge that accelerates innovation in government, industry, and society.

Kong is an example of the diverse collaborators in a CRT which comprise statisticians, computing scientists, neuroscientists, and biomedical engineers. He applies his expertise in statistical machine learning and neuroimaging to this project by looking at possible connections between diseases and particular genes, or how a given area of the brain may be related to mental disorders such as Alzheimer’s disease or ADHD.

“This project was born out of a desire to increase interest in research and training in statistical science for neuroimaging and imaging genetics in Canada” says Nathoo. “In recent years this general area has received a lot of attention and has been growing within the area of statistics, but this growth within Canada has been somewhat limited.” Nathoo and Kong’s CRT hopes to change this.

An example of their work involves a recent study (Nie et al., 2019) where this team used a statistical model of brain dynamics and related the connectivity of several key brain regions to genetics. Using this approach, they found a genetic signal on chromosome 11 that is associated with a potential path of information flow involving three brain regions, starting from the medial prefrontal cortex to the right intraparietal cortex going through the left intraparietal cortex. In a separate analysis, they also found that this same genetic signal is potentially associated with the probability of Alzheimer’s disease and had a relatively large effect relative to other genetic markers. This exciting result has motivated further studies with emphasis on replication and further study. These studies have the potential to further our understanding of Alzheimer’s disease, genetics and brain connectivity.

Collaborations involving neuroscience and statistics are widespread and have led to new innovations and insights in both fields. Nathoo and Kong envision a future where Canadian statisticians are heavily involved in this research. Together with statistician Bei Jiang and medical imaging researcher Faisal Beg, they are currently working on another exciting neuroimaging project involving the integration of computer models of the brain that generate synthetic fMRI data and statistical approaches for combining this computer model output with real fMRI data – so called data assimilation.  

CANSSI is providing funding for this neuroimaging project, with St. Joseph’s Healthcare, McMaster University Medical School, University of Victoria, the University of Alberta, and the Natural Sciences and Engineering Research Council also contributing.

Learn more about how CANSSI is providing the leadership to build multi-disciplinary collaborations across the statistical and data science fields: www.canssi.ca

Comments are closed.