Neurosalience #S4E18 with Vince Calhoun - (Part 2/2) A principled approach to data mining

Dr. Vince Calhoun is the founding director of the tri-institutional center for translational research in Neuroimaging and Data Science (TReNDS) which is a consortium formed by Georgia State University, Georgia Tech, and Emory University.

In this part 2 of Peter and Vince’s discussion, they dive further into addressing the challenges that fMRI and other modalities face in finding useful information about psychiatric disorders that can be used clinically. They talk about what neuroimaging has taught us about schizophrenia, as well as the goals and challenges of establishing clinical relevance. They also talk a bit about the importance of a data driven approach to development of processing methods, as well as variability in fMRI data, and the challenges and opportunities that big data sets offer, the promise of data fusion, and multivariate modeling. Lastly, they also discuss his latest work in deep learning and what it offers, and spend quite a bit of time discussing data driven approaches vs model driven approaches.

This discussion was an outstanding perspective builder. We hope that you enjoy it!

Episode producers:

Omer Faruk Gulban

Xuqian Michelle Li