03/30/2023: Bratislav Misic

[GabeNicholson]

Comment below with a well-developed question or comment about the reading for this week's workshop.

If you would really like to ask your question in person, please place two exclamation points before your question to signal that you really want to ask it.

Please post your question by Tuesday of the coming week, at 11:59 PM. We will also ask you all to upvote questions that you think were particularly good. There may be prizes for top question-askers.

[sdbaier]

Dear Professor Misic, thank you for sharing your research with us. Great to have some Canadian representation at our workshop! A few questions:

Functional Connectivity emerges from higher-order interactions, i.e., structure is linked to function via complex associations, but their correlation is imperfect at best. Estimating FC across time seems logical, but how can these associations be estimated across individuals (presumably) within a single time slice? Is the underlying assumption that neural activity in different individuals corresponds to different “stages”, analogous to estimation across time?

The paper presupposes structural connections, and indicates that functional connections emerge from higher-order interactions. How can we be sure about the directionality of this link? This is the sociologist in me, so this might not apply to neuroscience, but could structure equally emerge from function?

[yuzhouw313]

Hello, Professor Misic,
Thank you for sharing with us your research on the relationship between brain's structural and functional properties at the macro-scale level. I look forward hearing more of your research on Thursday!

I am not familiar with the field of neuroscience or neuro-imaging techniques, but I found your discussion on the relationship and non-one-to-one mapping between the structural and functional connections of the brain interesting and relatively intuitive for a layperson like me. However, I am particularly curious about the disruption or change on this relation, especially its consequences. Specifically, when you stated that "disruption of the protein’s structure results in loss of function," what function are you referring to and can you elaborate on the impact of brain's structure disruption on its function? Thank you again!

[secorey]

Hi Professor Misic,
In my limited exposure to neuroscience, I remember learning about how for some people with portions of their brain removed, other parts of their brain learn to execute the same functions that the missing piece used to control. Is this related to your research and, if so, how has it informed your knowledge of the nature of structure-function relationships?
Thanks for coming to present this paper.
Sam

[LynetteDang]

Hi Professor Misic, thank you for sharing your work with us. I have a couple of questions:

1. how do different types of brain activity (e.g. resting-state vs. task-based) relate to brain networks?
2. how do individual differences in SC and FC relate to differences in cognitive function, personality, and behavior?
3. what are the ethical implications of studying structural connectivity in the brain, particularly with regard to privacy and informed consent?

Thanks!

[bmisic]

Dear Professor Misic, thank you for sharing your research with us. Great to have some Canadian representation at our workshop! A few questions:

Functional Connectivity emerges from higher-order interactions, i.e., structure is linked to function via complex associations, but their correlation is imperfect at best. Estimating FC across time seems logical, but how can these associations be estimated across individuals (presumably) within a single time slice? Is the underlying assumption that neural activity in different individuals corresponds to different “stages”, analogous to estimation across time?

The paper presupposes structural connections, and indicates that functional connections emerge from higher-order interactions. How can we be sure about the directionality of this link? This is the sociologist in me, so this might not apply to neuroscience, but could structure equally emerge from function?

Great question! We've actually done just such a study, looking at how frame-by-frame fluctuations in functional co-activation relate to the underlying structural connectivity:

https://doi.org/10.1038/s42003-022-03466-x

More broadly, all the methods typically used to study structure-function coupling can be applied to individuals, and there is an emerging literature on how individual differences in structure-function coupling relate to behaviour, cognition, etc, and how they are affected by disease.

[bmisic]

Hi Professor Misic, In my limited exposure to neuroscience, I remember learning about how for some people with portions of their brain removed, other parts of their brain learn to execute the same functions that the missing piece used to control. Is this related to your research and, if so, how has it informed your knowledge of the nature of structure-function relationships? Thanks for coming to present this paper. Sam

There is indeed a perpetual interplay between structure and function, and this is why the notion of polysynaptic signaling in brain networks is so important. Namely, a complex network can support multiple functional configurations, so some perturbations can theoretically be compensated for by alternative signaling patterns.

[Thiyaghessan]

Dear Professor Misic,

Thank you for the talk! In your paper you briefly mention reservoir computing at the tail end:

"Particularly exciting are advances in artificial intelligence, such as reservoir computing, that make it possible to
implement connectomes as artificial neural networks"

I was wondering if you could elaborate on how reservoir computing will allow us to bridge the gap between neural networks as understood today in deep learning and biological neural networks in the human brain.

Thank you!

[kuitaiw]

Dear Prof Misic,
Neuroeconomics is a discipline that attempts to understand the neural mechanisms underlying economic behavior. So I would like to ask whether we can use techniques such as brain imaging to study the neural activity of individuals when making economic decisions? If so, what exactly should we do?

[borlasekn]

Prof. Misic - thank you for sharing your expertise! I appreciate hearing from perspectives outside of my own field and enjoying thinking about how work can translate between fields. That being said, as an individual interested in educational research, your section on "Individual Differences in Structure-Function Coupling" was particularly interesting. Individual-specific neuroscience is important in child development, especially in understanding how life events effect brain function during youth development! You mention that respecting individual differences is a major question for future research studies on the structure-function relationship. Would you mind giving some examples on how this research would be carried out, and what key insights you'd anticipate from this research? Thanks!

[taizeyu]

Dear Prof Misic,

Thank you for sharing your research. This is a very profound paper. I am curious that how can we apply the technology mentioned in the paper to the world. And in what areas can it be applied?

Thank you very much

[JoeHelbing]

Professor Misic,

I also have very limited exposure to Neuroscience, so the concept of functional vs structural connectivity is sort of hard to grasp. My best understanding from the paper was that SC is what neurons are connected to what neurons, whereas FC is what neurons work together with what neurons? Apologies is that is massively off-base. It's interesting that these would not be 1:1, as a non-neruo person, more SC more FC would make sense to me.

Could you give a bit more background for why this is an area of study? I'm making an assumption here that it is more "fundamental" science, laying foundations for more applied work?

[zoeyjiao1104]

Hi Professor Misic,

Thank you so much for sharing your work. I was curious about how you would address the ethical issues that could potentially arise in your study in terms of participant consent. In addition, could you please elaborate on the real world application of your study. Thank you!

[sushanz]

Dear Professor Misic, I appreciate your time and thanks for sharing! Really interesting work you have done and presented in the paper. As we are aware, deep learning has made remarkable progress in the field of cognitive science. People deploy today's techniques to study human behavior and brain function. You mentioned in your paper that the current models lack biological details therefore hinder the functional ability to fully emulate the brain and predict functions.However, I am curious to know why this field is relevant if we can never gather enough information to create a flawless model? Especially, there are many inexplicable and mental processes involved in the behavior we study, such as the availability heuristic, which has both biological and cognitive influences. Although without the biological details, some of the techniques today has already been able to replicate the process and taught our machine to recognize and recall relevant patterns

[edelahayeUChicago]

Dear Professor,

I come from an economics background so I won't pretend that I understand most of your paper's contents. I'm interested though what the key takeaways in terms of high level behaviour as might be analysed in other social and behavioural sciences? Is this something that you could see becoming a more fundamental direction of research in how social behaviour is determined?

Thanks,
Elliot

[awaidyasin]

Dear Prof., thanks for sharing your work! On page 306, you mention, "Perhaps the simplest way to link structure and function is statistically. Varying forms of reduced rank regression have emerged as particularly useful, including canonical correlation [52] and partial least squares". I was wondering how would the empirical evidence (i.e., the underlying dataset) be generated for such applications given that this area studies quite specific connections in the brain.