24. Brain structure-function association and complexity

 

Kaustubh Patil

We apply ML/AI techniques to neuroimaging data to understand brain organization and its brain-behavior relationships in health and disease. A current research direction is use of ML/AI method for understanding complexity of brain function (3D time series data) and its association with brain structure, e.g. using VAEs. the We also develop general purpose ML methods and tools.

What is the data science project’s research question? How can we design ML/AI methods to capture information in 3D time series data (functional MRI scans) that will help us understand how brain structure generates function and how it might be disturbed in disorders.

What data will be worked on?  MRI data from functional (time series) and structural (wiring) scans (at least two samples; N>1000 and N>10K)

What tasks will this project involve?  Data wrangling; Pipelines using traditional dimensionality reduction methods (e.g. PCA); Pipelines using state-of-the-art variational auto-encoders

What makes this project interesting to work on?  Large neuroimaging data; basic and applied scientific questions; mix of conventional and deep-learning methods; contribute towards open-source tools

What is the expected outcome?   Contribution to research paper, Contribution to software development

What infrastructure, programs and tools will be used? Can they be used remotely?   All the infrastructure can be used remotely. We have a dedicated HPC cluster which can be used remotely: with 1,300+ CPU cores, 80 GPUs (72x Tesla P100, 8x GTX 1080Ti), and 16+ TiB RAM. We also have access to the Juelich Supercomputers.

What skills are necessary for this project? Data analytics / statistics, Data mining / Machine learning, Deep learning

Is the data open source?   Both open-source and upon request data will be used

Interested candidates could be at master, PhD and post-doc levels. Kaustubh Patil is looking for 2 visiting scientists, working on the project together with the team.