Santa Fe Institute Collaboration Platform

COMPLEX TIME: Adaptation, Aging, & Arrow of Time

Get Involved!
Contact: Caitlin Lorraine McShea, Program Manager,

Cognitive Regime Shift I - When the Brain Breaks

From Complex Time

Category: Application Area Application Area: Aging Brain

Date/Time: July 23, 2018 - July 25, 2018

Location: Santa Fe Institute (Collins Conference Room)

Upload a group photo


  • Steven Petersen (Washington Univ.-St. Louis)

  • Audio files

    Upload an audio file


    Click each agenda item's title for more information.
    Monday, July 23, 2018
    9:00 am - 9:30 am Day 1 Continental Breakfast (outside SFI Collins Conference Room) Download Presentation (Encrypted)
    9:30 am - 10:00 am Welcome & Introduction around the Room - David Krakauer (SFI), Susan Fitzpatrick (JSMF), Steven Petersen (Washington Univ.-St. Louis)
    10:00 am - 10:30 am WG Context under SFI Adaptation, Aging, Arrow of Time (AAA) & Wiki Collaboration Platform - Amy P Chen (SFI) Download Presentation
    10:30 am - 11:20 am "The use of large-scale brain correlations to study aging and some interesting issues that they raise" - Steven Petersen (Washington Univ.-St. Louis)
    11:20 am - 12:10 pm Consciousness, Cognition, and the Prefrontal Cortex - George Mashour (Univ. Michigan)
    12:10 pm - 1:00 pm Day 1 Lunch (outside SFI Collins Conference Room)
    1:30 pm - 2:20 pm "Prion dynamics and latency" - David Krakauer (SFI)
    2:20 pm - 3:10 pm Neuronal Avalanches - Dietmar Plenz (NIH)
    3:10 pm - 3:30 pm Day 1 PM Break
    3:30 pm - 4:20 pm Network Breakdown Phenomena - Sidney Redner (SFI) Download Presentation
    4:20 pm - 5:10 am On the Stability of Large Ecological Communities - Jacopo Grilli (ICTP) Download Presentation
    6:30 pm Day 1 Group Dinner at La Boca
    Tuesday, July 24, 2018
    9:00 am - 9:30 am Day 2 Continental Breakfast (outside SFI Collins Conference Room)
    9:30 am - 10:00 am Recap from Day 1
    10:00 am - 10:50 am How Does Context Impact Cortical Development - Leah Krubitzer (UC Davis) Download Presentation
    10:50 am - 11:40 am Large-scale Brain Network Changes Across the Healthy Adult Human Lifespan: Relations to Cognition and First Steps toward Identifying Potential Risk Factors of Brain Decline - Gagan Wig (UT Dallas)
    11:40 am - 12:30 pm Day 2 Lunch (outside SFI Collins Conference Room)
    12:30 pm - 1:20 pm The Brain and other Networks - Graham H. Creasey (Stanford)
    1:20 pm - 2:10 pm States and Stability in Human Functional Brain Networks - Caterina Gratton (Northwestern Univ.) Download Presentation
    2:10 pm - 2:30 pm Day 2 PM Break
    2:30 pm - 3:20 pm Gene Networks in Brain and Neurodegenerative Disorders - Daniel Geschwind (UCLA)
    3:20 pm - 5:00 pm Open discussion, synthesis, planning for Day 3, platform time
    6:30 pm Day 2 Dinner: self-organize
    Wednesday, July 25, 2018
    9:00 am - 9:30 am Day 3 Continental Breakfast (outside SFI Collins Conference Room)
    9:30 am - 12:00 pm Research Jam
    12:00 pm - 1:00 pm Day 3 Lunch (outside SFI Collins Conference Room); Adjourn

    Add an Agenda Item[edit source]

    Meeting Synopsis

    In this inaugural working group on Cognitive Regime Shift, complexity tools and models developed to understand network phenomena in physiological, sociological, ecological, and financial realms will be applied to better understand the loss of cognitive function. The central goal is to achieve synthesis and integration between neurology and several areas of complexity science bearing on network failure, such that interventions for increasing long-term adaptability of the brain might be found. Critical questions that will be raised during the working group include relationship between network failure, adaptation and robustness to system aging, early indicators for risk factors, application of criticality, long-range order, and collective dynamic to the aging brain, and specific aspects of the above to Alzheimer’s.

    Some questions that will be discussed during this working group are:

    1. What is the relationship between network, failure, adaptation, and robustness to system aging?

    2. What are the early indicators and risk factors that are predictive of loss of functions?

    3. Can ideas from the study of criticality, long-range order, and collective dynamics inform our understanding and suggest restorative intervention in the aging brain?

    4. How can neurodegenerative disease such as Alzheimer’s be better characterized as atrophy on large-scale, distributed function-critical neural networks?

    Additional Meeting Information

    All talks are 30 minutes, followed by 20 minutes of open discussion

    Abstracts by Presenters

    David Krakauer (SFI) - "Prion dynamics and latency"[edit source]

    A large number of neurodegenerative diseases feature the accumulation of mis-folded proteins. These include prion diseases, Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease. In all of these cases several different scales of organization are associated with disease progression or onset to include genetic, epigenetic, neural circuits, brain modules, and behavior. How should we best integrate data from each of these levels and what models and theories allow us to span levels? I shall discuss a few dynamical models of polymerization, protein accumulation, and protein diffusion through neural connections, that provide insights into disease progression at a number of different time and space scales. An ongoing challenge is a criterion for fixing thresholds that define an observable cognitive regime shift.

    George Mashour (Univ. Michigan) - Consciousness, Cognition, and the Prefrontal Cortex[edit source]

    The goal of my talk was to inform the group on the current controversies in identifying the neural correlates of consciousness, to distinguish phenomenal vs. access consciousness, and to highlight the potential role of recurrent processing (also known as reafferent, reentrant, reverberant or feedback processing) in conscious experience. I also wanted to show that anesthesia can be a reversible, functional model of a "broken brain" and demonstrate interventions (in this case, cholinergic stimulation of prefrontal cortex) that might reverse phenotypes of brokenness.

    Gagan Wig (UT Dallas) - Large-scale Brain Network Changes Across the Healthy Adult Human Lifespan: Relations to Cognition and First Steps toward Identifying Potential Risk Factors of Brain Decline[edit source]

    -Human brain areas are organized into a large-scale functional network, which can be measured at rest using non-invasive brain imaging (functional MRI)

    -The brain network contains segregated sub-networks that correspond to functionally specialized brain systems

    -The segregation of brain systems declines with increasing age, across the healthy adult lifespan

    -System segregation relates to cognitive function in individuals (greater system segregation is associated with better long--term memory ability)

    -Certain health risk factors (e.g., lower socioeconomic status) are related to lower system segregation

    -My working hypothesis is that gradual and sudden cognitive decline is related to changes in system segregation as an individual ages, and that individual differences in rate and risk of decline are a consequence of the capacity of the functional brain network to tolerate and adapt to damage (neurodegeneration)

    Sidney Redner (SFI) - Network Breakdown Phenomena[edit source]

    I gave a basic review of percolation theory on lattices and outlined the behavior of physical observables, such as the correlation length, the mean cluster size, and the percolation probability on the bond occupation probability. I then discussed the analogous percolation transition on complex networks, where the degree distribution can be broad. The basic new feature of complex networks is that they are relatively robust to random removal of nodes or links and quite vulnerable to the removal of the highest-degree nodes.

    Finally, I presented two examples of network breakdown phenomena: the electrical failure of electrical networks of fuse elements and the external voltage is increased, and the clogging of fluid networks during the process of filtration.

    Dietmar Plenz (NIH) - Neuronal Avalanches[edit source]

    Jacopo Grilli (ICTP) - On the Stability of Large Ecological Communities[edit source]

    Ecological communities (more generally, non-linear systems) often showmultiple regimes, which are separated by a sharp and rapid transition. I will discuss the scenario when the driver of the transition is the structure of interactions. Random matrix theory has a powerful set of tools that can be used to unveil the relation between interaction structure and dynamics.

    Take home messages:

    - universality: when many components interact many details do not matter (e.g. the distribution of interaction coefficients) and few global properties of the interactions determine the relevant dynamical properties - the effect of the structure (whether a given network structure is stabilizing or destabilizing compared to the null/random case) *depends* on the interaction strengths properties

    Caterina Gratton (Northwestern Univ.) - States and Stability in Human Functional Brain Networks[edit source]

    In my talk, I went over two main empirical studies. In the first, we used a dataset that highly-sampled 10 individual subjects, across different days and tasks. We asked how functional brain networks vary over different timescales, and found that these network measurements are primarily stable, with only moderate/minor state-based effects. In the second experiment, we looked at how Parkinson’s disease affect functional brain networks. We found that PD selectively impacts blocks of network-to-network connections, remote from primary pathophysiology. I also described initial findings from a recent initiative into how we might characterize individual variation in brain networks, showing that individual network “variants” are stable and systematic. From these findings, I concluded:

    1. Functional network measures are well-suited to tracking slow, stable brain processes
    2. These measures can provide detailed images of individual differences
    3. Functional network effects can be complex, occurring at locations remote from primary pathology

    Graham H. Creasey (Stanford) - The Brain and other Networks[edit source]

    The brain, whether considered as a network or a complex adaptive system, is obviously linked to other networks and complex adaptive systems, both technical and social. These links are usually mediated by inputs and outputs corresponding to sensors (visual, auditory, tactile, etc) and actuators (muscles, glands, etc) but it is also possible to create direct electrical interfaces to the nervous system. Study of these inputs and outputs gives insights into the internal function of the brain as a network.

    How do the brain and other networks adapt/learn/grow in parallel or collaboratively? How can this knowledge be used to defer or prevent network failure, especially with aging?

    Post-meeting Summary by Organizer[edit source]

    It could be argued that the brain is the prototypical complex system: in humans approximately 86 billion nerve cells are coordinated into tens of cortical regions to ensure that information from sensation and perception is utilized in a repertoire of adaptive behaviors. And just as the information processing principles that allow a large decentralized system to function presents many new scientific challenges, so does the unique way in which the brain fails, either through normal aging or through the vastly accelerated loss of function associated with neurodegenerative disease. This meeting sought to employ insights of complexity science as they pertain to system collapse and system robustness to new emerging data accumulated by neurocognitive researchers. Of particular interest to the group was the question of network failure, or cascading failure. These describe the accelerating loss of modules, units, and cells associated with the reduction or sparsification in network connectivity observed in a range of diseases from Alzheimer’s through to prion diseases. The group focused on new and potential data sets and questions related to their optimal spatio-temporal resolution as well as new insights from nonlinear dynamics to include the statistical physics of networks and the stability of analogous ecological networks applied to clinical and psychological observation. Throughout discussion much emphasis was placed on the discovery of the appropriate levels for causal explanation. Candidate levels included the genetic level, protein aggregations within a single nerve cells, local neural circuits, and large-scale cortical fields. At each level there exists data associated with neuro- degenerative phenotypes.

    At this point it is not clear what is purely correlational versus what could be said to be truly causal. In the absence of novel experimental procedures or realistic computer simulations the field is dominated by correlational studies and descriptions. Evidence from the loss of key brain function under anesthetic provided correlational evidence for critical regions and possibly pathways that might recapitulate the loss of conscious states of awareness during senescence. The loss of coherent correlations in phase among distant cortical regions in the resting state are reported to be correlated with the loss of key cognitive capacities in a variety of stressful circumstances, including sustained poverty. Evidence from the comparative analysis of cortical fields in nonhuman vertebrates suggests significant variability in those pathways required to ensure continued cognitive function over a lifespan. In many cases where data can be obtained there is evidence for the statistics of neural avalanches or power-law like behavior in the distribution of brain activity, suggesting that deviations from baseline power-law distributions might be profitably used as a diagnostic tool in establishing the effective state of health of the nervous system. At this point the group feels the need for a systematic summary of the key insights of network science as they pertain to network collapse, combined with a deliberative effort to map the requirements of these models onto available data-sets at different levels of organization to include the genetic, neural, neural circuit, and cortical field.

    It was widely agreed that there is value to the community of aging and brain disease researchers to becoming more familiar with techniques that are a good fit to the essential characteristics of the system that they are working with. The group intends to spawn a small number of distinct groups each of which will pursue particular neurodegenerative diseases using a variety of new techniques.

    Additional Post-meeting Summary by Organizer
    What was the big question or idea the meeting was designed to explore?
    Coming soon.
    What was the goal of the meeting?
    Coming soon.
    What was the single most important outcome of the meeting?
    Coming soon.
    What disciplines/fields were represented at the meeting?
    Coming soon.
    What one or two new research direction(s) did the meeting suggest?
    Coming soon.
    What was the most interesting thing said during the meeting, and who said it?
    Coming soon.
    What idea from the meeting is likely to be most impactful for science?
    Coming soon.
    What idea from the meeting is potentially translatable into an application?
    Coming soon.
    What new method/technology/algorithm/etc. resulted (or may result) from this meeting?
    Coming soon.

    A narrative summary of the meeting

    Coming soon.

    Outcomes that might emerge from the meeting

    Coming soon.

    Communication produced related to the meeting

    Coming soon.

    Post-meeting Reflection by Presenter

    David Krakauer (SFI) - "Prion dynamics and latency" Link to the source page[edit source]

    A number of critical questions were raised about the best levels at which to establish causality when it comes to understanding both natural and disease-related aging. Namely what are the best observables to consider? Should these be single measurements or network based measurements. Could the best indicators involve comparisons across genetic and cognitive networks applying similar methods, or as is more typical time-dependent changes in a given network at one level of analysis. A recurring question was the relationship between energy and information and how their reciprocal dependencies change over the course of time and the course of disease.

    Some very general issues that arose in conversation that require further exploration include:

    1. Approaching disease from a first-principles theoretical perspective - as is common in ecology - thus establishing principled data collection objectives (this would require a rigorous operational definition of the disease state in formal terms)
    2. The value and limitation of the current inductive, big data approach, that focuses on time-dependent associations
    3. The meaning of cognitive reserve, exercise or error correction, and the limits to these
    4. How adaptive phenomena that are ongoing mitigate the disease state or at some point perhaps accelerate it.
    5. How we might better explore causality in large systems with extensive non-linear feedback mechanisms.
    log in to comment

    Jacopo Grilli (ICTP) - On the Stability of Large Ecological Communities Link to the source page[edit source]

    - dynamics: if we think at aging as the approach of critical transition we have specific predictions: critical slowing down, flickering, etc

    - information: where is information in this dynamical / tipping point picture? Do we need it?

    - scales and information: at what scale should we look at the brain to "explain" the interesting (information) phenomena?

    log in to comment

    Sidney Redner (SFI) - Network Breakdown Phenomena Link to the source page[edit source]

    I'm interested in datasets that could serve as a diagnostic for failure of brain networks. It was mentioned at the meeting that there exists data for response time as a function of age. Perhaps this could be used to understand the performance of the brain network as a function of age.

    More general question: is there an unambiguous way to determine age by measuring some aspect of brain function?

    log in to comment

    Graham H. Creasey (Stanford) - The Brain and other Networks Link to the source page[edit source]

    Recognition of depth of knowledge in related disciplines/communities/clusters.

    Recognition of how little we connect between disciplines/communities/clusters.

    What disciplines/people/agents are missing from our meeting? What important elements are missing from our model of the system needed to understand/influence failure of the brain network?

    How do we maintain communication between these agents/communities in between physical or virtual meetings?

    log in to comment

    Gagan Wig (UT Dallas) - Large-scale Brain Network Changes Across the Healthy Adult Human Lifespan: Relations to Cognition and First Steps toward Identifying Potential Risk Factors of Brain Decline Link to the source page[edit source]

    -applying different approaches in complexity towards understanding age-related changes in brain organization (e.g., thinking about flickering, network 'clogging')

    -prospects of expanding/linking our work to additional scales of analysis

    log in to comment

    Caterina Gratton (Northwestern Univ.) - States and Stability in Human Functional Brain Networks Link to the source page[edit source]

    One discussion that emerged after the talk was the question of what the right level of analysis was for understanding brain dysfunction in aging and PD, and what form of causal argument or mechanism can be derived from these network descriptions of brain organization and dysfunction. A very interesting direction to go would be to create more theoretically driven models of brain dysfunction in PD, that might explain the disconnect between the functional network effects and known pathology in the disease. These models could then be tested in future experiments.

    log in to comment

    George Mashour (Univ. Michigan) - Consciousness, Cognition, and the Prefrontal Cortex Link to the source page[edit source]

    The meeting was illuminating in a number of regards. In addition to the new content knowledge, the framework of emergence/causality/time/complexity was of great interest and utility. In terms of specific knowledge that will inform my future work, the limitations of DTI as a metric for human structural connectivity was important to learn. Also, the lecture on critical dynamics was- in my opinion- important in linking scales from neuronal spike activity to large scale networks. Criticalitycan potentiallyfunction as a surrogate for optimal "health" in the system and distance from criticality can potentially function as a surrogate for "disease."

    log in to comment
    Post-meeting Reflection by Non-presenting Attendees

    Reference Materials by Presenting Attendees[edit source]

    Gagan Wig (UT Dallas) - Large-scale Brain Network Changes Across the Healthy Adult Human Lifespan: Relations to Cognition and First Steps toward Identifying Potential Risk Factors of Brain Decline[edit source]

    Title Author name Source name Year Citation count From Scopus. Refreshed every 5 days. Page views Related file
    Decreased segregation of brain systems across the healthy adult lifespan Micaela Y. Chan, Denise C. Park, Neil K. Savalia, Steven E. Petersen, Gagan S. Wig Proceedings of the National Academy of Sciences 2014 0 5

    Caterina Gratton (Northwestern Univ.) - States and Stability in Human Functional Brain Networks[edit source]

    Here are some references to our work that I discussed which could be relevant:

    Title Author name Source name Year Citation count From Scopus. Refreshed every 5 days. Page views Related file
    Precision Functional Mapping of Individual Human Brains Evan M. Gordon, Timothy O. Laumann, Adrian W. Gilmore, Dillan J. Newbold, Deanna J. Greene, Jeffrey J. Berg, Mario Ortega, Catherine Hoyt-Drazen, Caterina Gratton, Haoxin Sun, Jacqueline M. Hampton, Rebecca S. Coalson, Annie L. Nguyen, Kathleen B. McDermott, Joshua S. Shimony, Abraham Z. Snyder, Bradley L. Schlaggar, Steven E. Petersen, Steven M. Nelson, Nico U.F. Dosenbach Neuron 2017 320 8
    Reference Materials by Non-presenting Attendees

    General Meeting Reference Material[edit source]

    Title Author name Source name Year Citation count From Scopus. Refreshed every 5 days. Page views Related file
    Aging and measures of processing speed Timothy A. Salthouse Biological Psychology 2000 672 59
    Open questions in artificial life Mark A. Bedau, John S. McCaskill, Norman H. Packard, Steen Rasmussen, Chris Adami, David G. Green, Takashi Ikegami, Kunihiko Kaneko, Thomas S. Ray Artificial Life 2000 159 17
    High performance communication by people with paralysis using an intracortical brain-computer interface Chethan Pandarinath, Paul Nuyujukian, Christine H. Blabe, Brittany L. Sorice, Jad Saab, Francis R. Willett, Leigh R. Hochberg, Krishna V. Shenoy, Jaimie M. Henderson eLife 2017 152 6
    Review - Segregated Systems of Human Brain Networks Gagan S. Wig Trends in Cognitive Sciences 2017 89 7
    Critical networks exhibit maximal information diversity in structure-dynamics relationships Matti Nykter, Nathan D. Price, Antti Larjo, Tommi Aho, Stuart A. Kauffman, Olli Yli-Harja, Ilya Shmulevich Physical Review Letters 2008 66 11
    Irregular spiking of pyramidal neurons organizes as scale-invariant neuronal avalanches in the awake state Timothy Bellay, Andreas Klaus, Saurav Seshadri, Dietmar Plenz eLife 2015 63 9
    Predicting the stability of large structured food webs Stefano Allesina, Jacopo Grilli, György Barabás, Si Tang, Johnatan Aljadeff, Amos Maritan Nature Communications 2015 57 8
    Coordinated reset vibrotactile stimulation shows prolonged improvement in Parkinson's disease Judy Syrkin-Nikolau, Raumin Neuville, Johanna O'Day, Chioma Anidi, Mandy Miller Koop, Talora Martin, Peter A. Tass, Helen Bronte-Stewart Movement Disorders 2018 11 15
    Voltage imaging of waking mouse cortex reveals emergence of critical neuronal dynamics 0 13
    Socioeconomic status moderates age-related differences in the brain’s functional network organization and anatomy across the adult lifespan 0 10
    Editorial overview: Neurobiology of cognitive behavior: Complexity of neural computation and cognition Alla Karpova, Roozbeh Kiani Current Opinion in Neurobiology 2016 0 5
    Loss of Consciousness Is Associated with Stabilization of Cortical Activity2 0 3