Michael W. Cole

Professor at Rutgers, director of a cognitive & network neuroscience lab – neuroimaging, connectivity, cognitive control, NeuroAI. Author of Brain Flows

[Michael W. Cole @TheColeLab]

Inspired by Edinburgh’s legacy of big ideas (e.g., Hume), on a train leaving the city, I crafted a book proposal about big ideas in network neuroscience—how brain flows generate representations that compose our minds. The rest is history, and this book: https://t.co/EwWRS1OIK0

[Michael W. Cole @TheColeLab]

@vineettiruvadi I disagree. I did cite single cell work and integrate it into the Brain Flows framework. There are many levels of organization in brains and the cellular level is an important one

[The Transmitter @_TheTransmitter]

In his review of Michael W. Cole’s @TheColeLab new book, “Brain Flows,” Daniel Graham highlights how Cole makes a compelling case for why dynamic movement of brain signals matters, despite missing mention of relevant theoretical and single-cell work.
https://t.co/G1jtuYjCs9

[Michael W. Cole @TheColeLab]

Daniel Graham gives an honest assessment of my just-published book (Brain Flows). Check it out!

He mentions researchers I underemphasized. The book stands on the shoulders of giants, and it’s unfortunate I didn’t have room to acknowledge them all!

https://t.co/JxM2cqLWnG

[Michael W. Cole @TheColeLab]

@DAgourakis In our tests (so far) rest FC approaches task FC but doesn't close the gap. It'd be interesting if it did in some cases, though.

[Michael W. Cole @TheColeLab]

I wrote an essay for Princeton University Press's Ideas! The essay dovetails with the core ideas in my just-published book "Brain Flows: How Network Dynamics Compose the Human Mind"

The essay:
What Brains Are For
https://t.co/dmwafQaBSc

[Michael W. Cole @TheColeLab]

@DAgourakis @DarwinCompPsych The FC is second-order (covariance), so it shouldn't be a problem. The mean activations removed from the FC are the exact values that are being predicted. I used simulations and empirical tests to double check this, here: https://t.co/wauCbmT6vm

[Michael W. Cole @TheColeLab]

@DarwinCompPsych @DAgourakis Yes, that'd probably be cleaner in some sense, but it'd also be conservative in the sense that task FC estimates should already have the mean task-evoked activity removed. So, it'd be a kind of "double cross-validation" (removing mean task-evoked variance in 2 ways).

[Michael W. Cole @TheColeLab]

@DAgourakis Note that the 0.13 value was lower than normal because I restricted the amount of data for resting-state FC estimates. This was to match the amount of task data I had per condition with HCP data. Later analyses in that paper show that rest FC approaches task FC with more data.

[Michael W. Cole @TheColeLab]

@SaintTJ_NUC Yes, the framework characterizes goal selection in terms of core values at the top of the brain's global flow hierarchy. All manner of strategies can be used to pursue goals, since goal pursuits are anchored in core values.

[Michael W. Cole @TheColeLab]

@DAgourakis Interesting. I haven't estimated the source of shared noise directly. Figure 9 of this study: https://t.co/zhdYppR63l has some relevant information. It shows that individual differences in task-specific functional connectivity plays a big role.

[Michael W. Cole @TheColeLab]

@DAgourakis They can be estimated as the multiplication of source neural activity multiplied by the source-to-target connectivity. Accuracy of the flow estimate can be determined based on prediction accuracy of the (held-out) target activity level.

[Michael W. Cole @TheColeLab]

@DAgourakis The brain flow patterns themselves are the computational primitives. A flow is the causal impact of one neural population on another, mediated by aggregate action potentials from the source and neurotransmitter release at the target.

[Michael W. Cole @TheColeLab]

I was interviewed at Rutgers University about my just-published book "Brain Flows: How Network Dynamics Compose the Human Mind" (https://t.co/S6IyeXG9SH)

The interview:
https://t.co/ehmZwkzvrA

[The Transmitter @_TheTransmitter]

In this episode of Brain Inspired, @TheColeLab explains how "brain flows" are the fundamental units of brain function and cognition.
https://t.co/Of35u11zcL

[The Transmitter @_TheTransmitter]

In his new book, @TheColeLab examines how the movement of neuronal activity between neuronal populations shapes just about everything the brain does.
https://t.co/NGAMV3FFwz

[Michael W. Cole @TheColeLab]

My latest Neural Flows post!

"Emotions as global low-dimensional brain flow patterns"

Based on the book Brain Flows: How Network Dynamics Compose the Human Mind

https://t.co/ruhoPMOYT8

[Michael W. Cole @TheColeLab]

Thanks to first author Lakshman Chakravarthula
and coauthors Takuya Ito and Alexandros Tzalavras!

The paper: https://t.co/OC24mkmLvb
https://t.co/bT6uJPepzD
19/n

[Michael W. Cole @TheColeLab]

Do brain connections just carry information between regions, or do they transform it?

Lab's latest out in Nature Neuroscience, led by Lakshman Chakravarthula: "Network geometry shapes the transformation of task representations across the human cortex"
https://t.co/bT6uJPdRK5 1/n

[Michael W. Cole @TheColeLab]

Together, these results suggest the brain's network geometry is a computational substrate, not just wiring: low-dimensional connectivity compresses representations for generalization, while high-dimensional connectivity expands them for task-specific implementation 18/n