OHBM Neurosalience OHBM

This episode’s guest is arguably one of the most influential scientists in the human brain mapping community. Dr. Peter Fox, director of the Research Imaging Institute at the University of Texas Health, San Antonio. Early in his career he wrote the seminal paper that showed, using positron emission tomography , that brain-activation related increases in blood flow are accompanied by only small increases in oxidative metabolic -  resulting in the blood locally increasing in oxygenation. This paper set the foundation for understanding all of Blood Oxygen Level Dependent Contrast used in fMRI today. The true purpose of activation-related flow increases is still an open question. The story of the events and details surrounding this are in his review article from the 2012 NeuroImage special issue. It's titled, simply "The coupling controversy."

Dr. Fox was also among the first to promote data sharing and pooling with his brainmap database, and early on, established stereotactic coordinates and spatial normalization as a way to put data into a shareable space. He started the annual meeting that pre-dated the Organization for Human Brain Mapping, and also founded one of the major brain mapping journals today, titled: Human Brain Mapping.

 

Peter had his formative undergraduate education at the extremely unique St. Johns college in Annapolis. He received his MD from Georgetown University, interned at Duke University, then carried out his residency and fellowship at Washington University where he worked closely with Dr. Mark Raichle, who was at the time pioneering PET scanning.

 

In this discussion, we delve into his contributions in a wide range of topics, from neurovascular coupling to the challenge of spatial normalization - particularly at high resolution - to subject variability, to clinical applications and the ongoing evolution of scientific publishing. Lots of history, content, and insight here. We hope you enjoy it!



Notable paper:

Fox PT., The Coupling Controversy, Neuroimage. 2012 Aug 15; 62(2): 594–601. 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4019339/



Episode producers:

Omer Faruk Gulban

Stephania Assimopoulos

In this episode, our guest is Rotem Botvinik-Nezer, a postdoc at Dartmouth University, working with Dr. Tor Wager in his  Cognitive and Affective Neuroscience Lab. In 2020, Dr. Botvinik-Nezer was first author of an influential paper published in Nature, titled Variability in the analysis of a single neuroimaging dataset by many teams, where the results were compared  from 70 independent teams analyzing a single data set having 9 hypotheses. This paper made it clear that there are many points of variability in data analysis pipelines, and provided further incentives for sharing data and code to grow consensus and replicability. While the popular press suggested that this paper was yet another hit to fMRI, we discuss how even papers that critique the results of this seminal paper ultimately converge in agreement with the overall message of systematic transparency. Dr. Botvinik-Nezer also has a strong interest in how our brains influence our perception of pain, having just published a recent paper showing evidence that regions associated with painful stimuli remain active even when subjects experience less pain while having the belief that a placebo is effective.

In this conversation, Peter and Rotem delve into all these topics and more, but spend  the bulk of the discussion on the interplay between choices in analyses, such as determining a statistical threshold, and  variability in results. We also discuss incentives for users to share data and code and possible ways to create a more solid scaffolding for best practices. 

Episode producers:

Omer Faruk Gulban

Xuqian Michelle Li

Dr. Daniele Marinazzo is a full professor in the department of data analysis at the University of Ghent, in Belgium. For over a decade he has been showing us what further information and insight we may extract from brain imaging data - from EEG and MEG to fMRI. He is technically a statistical physicist, but in reality, he is a network neuroscientist and data modeler who is constantly pushing the envelope.

In this podcast he discusses some recent papers that go into how we might be able to improve the impact and relevance of new findings and models through careful benchmarking and well considered experimental design.

He talks about his desire to move from correlation to causation in functional connectivity studies, he discusses granger causality, as well as moving from pairwise correlation to multivariate correlation.

Furthermore, he delves into the limits of hemodynamics - limits that may be pushed back to a degree, as suggested by his compelling work showing that hemodynamic response function, which varies over space, may be estimated on a voxel-wise basis using resting state data alone.

His work in estimating and mapping the Excitation/Inhibition ratio in the brain by using gamma frequency coherence as a signature was also discussed. This has potentially profound clinical and research applications.

Lastly, his collaborative work with the European Human Brain Project towards the creation of the useful website, called ebrains (https://www.ebrains.eu), was discussed, which serves as a repository and tool for exploring shared data and code, as well as providing a user-friendly encapsulation of the project's collective effort.

It is an all-around fun, eye-opening discussion featuring an outstanding scientist who is not only deep in the trenches of network modelling, but also a strong proponent of open science and constant engagement across disciplines.



Episode producers:

Omer Faruk Gulban

Alfie Wearn

Stephania Assimopoulos



Referenced Papers:

Mika Rubinov. Circular and unified analysis in network neuroscience. eLife. 2023; 12:e79559. Doi: 10.7554/eLife.79559

 

Reid AT, et al. Advancing functional connectivity research from association to causation. Nat Neurosci. 2019 Nov;22(11):1751-1760. Doi: 10.1038/s41593-019-0510-4.

 

Valdes-Sosa PA et al. Effective connectivity: Influence, causality and biophysical modelling. Neuroimage. 2009; 58(2): 339-361. Doi: 10.1016/j.neuroimage.2011.03.058.

 

Wu GR, et al. A blind deconvolution approach to recover effective connectivity brain networks from resting state fMRI data. Medical Image Analysis. 2013; 17(3):365-374. Doi:

10.1016/j.media.2013.01.003.

Today, we are excited to have Dr. Gang Chen on the podcast. Dr. Chen is the go-to statistics guru for the fMRI community at the NIH and a well-respected scientist worldwide. He is a staff scientist in the group that developed the AFNI software package. As an applied mathematician, Dr. Chen has written a series of insightful papers in the past seven years, bucking the status quo in fMRI processing - essentially saying that we are throwing away too much valuable information by thresholding our data, relying on overly simple and rigid models of the hemodynamic response, not mapping effect sizes, and using center of mass measures to describe clusters of activation. He backs it all up with a rigorous approach characterized by all good statisticians. He is a master in the art of casting a wide net to capture useful data without taking in artifact and noise, finding that sweet spot in data reduction to balance utility with sensitivity. 

In this episode, we hear all about Dr. Chen’s perspectives through these papers, which are so important yet not widely known or embraced by the field. We hope you enjoy it!

Episode producers:

Omer Faruk Gulban

Xuqian Michelle Li

In this episode, it's our pleasure to host Jack Wells who is a Wellcome Trust Career Development Fellow at the University College London Center for Advanced Biomedical Engineering. Dr. Wells received his Ph.D. in MRI in the Department of Medical Physics and Bioengineering at  at University College London in 2010, and since he began his scientific career, he's been working at the interface of MRI methodology and neurophysiology - focusing on understanding the Cerebral Spinal Fluid dynamics and how they may relate to the Glymphatic system. He and his colleagues have been among the leaders in using MRI to image and characterize the glymphatic system as well as the  brain - cerebrospinal fluid barrier. The glymphatic system is hypothesized to be the paravascular mechanism by which CSF is washed through brain tissue - typically during sleep - clearing out metabolic waste. It is an incompletely understood yet potentially profoundly important system where its dysfunction  may be at the root of  disorders that include Alzheimer's disease.

In this wonderful conversation we hear all about Jack's and others' work imaging and understanding the hydrodynamics and spatial organization of neurofluids in the brain.

We hope you enjoy this conversation as much as we did!

If you are interested in working with Nathan, he is currently recruiting for a postdoc! Send your CV to lbc.spreng@gmail.com



Today our guest is Nathan Spreng. Dr. Spreng is the James McGill Professor of Neurology and Neurosurgery and Director of the Laboratory of Brain and Cognition at the Montreal Neurological Institute and Hospital at McGill University.

As an undergraduate, Dr. Spreng was initially interested in pursuing a major in poetry until he took a psychology class that sparked his interest in the brain. He received Ph.D. in 2008 from the University of Toronto in Brian Levine's lab, and post docs with Cheryl Grady at the University of Toronto and Dan Schacter at Harvard. After about 5 years as an assistant professor at Cornell University, he moved to McGill University.

Throughout his career Dr. Spreng has been using fMRI to reveal subtle yet repeatable large-scale brain networks as they relate attention, memory, cognitive control, and social cognition. He has also helped to elucidate the central role that the default network plays in self-generated thought, and in how it dynamically interacts with multiple systems in the brain.

In this episode Peter and Nathan have a far reaching conversation about his work and what it implies, covering his study of age dependence of resting state hippocampal-linked network ensembles, how to move from mapping networks to modeling and understanding mechanisms, the many possible clinical implications of his work, current understanding of Alzheimer's disease, our mutual appreciation for multi-echo EPI, his data release paper of a large multi-echo EPI and structural MRI data set, and much more.

Enjoy listening!



Episode producers:

Alfie Wearn

Omer Faruk Gulban