Pre-OHBM Gradients Workshop

Session 1: Methods

Superficial white matter

  • Underneath grey matter, has been described in anatomical studies
  • SWM in development and aging and disease, temporal lobe epilepsy, AD, and ASD
  • Current methods to study SWM: surface based tractography, mask based methods (Registration-dependent), laplacian
    • disadvantages, computationally expensive, difficult to implmeent, multiple dependencies
    • new method to identify SWM
    • 7T MRI dataset - high res data, 3 sessions etc.
    • solves laplacian field, generates surfaces below grey matter (can be many surfaces)
  • microstructural intensity profiles derived from each SWM surface

Biologically annotated brain connectomes

  • connectomes: network of nodes and edges
  • annotated connectomes: thus far, pairwise relationships between regions
  • homophily in annotated connectomes: nodes that have degree in a connectome tend to be more connected
    • homophily can be quantified using the assortativity coefficient: the correlation between annotation values of connected nodes
  • spatial constraints of the brain: annotations are spatially autocorrelated and connections have a wiring cost (most connections are short range)
  • what is the relationship between local biological annotations and brain connectivity?
    • paths: what bio attributes do signals encounter along communication paths?
    • can biological annotations enhance models of the brain?
  • annotation-enhanced models of the brain:
    • biophysical models
    • heterogenous models
  • models of brain diseases
  • neuromaps toolbox: hosts brainmaps from lots of people etc.

Principles of white matter organization in relation to cognition in youth

  • Speaker from Ted Satterthwaite’s group
  • surface gradients: lower order sensory processing - complex cognitive functions
  • braod categories of gradients: function - structural: microstructure (myelination) - development:fluctuation amplitude (age effect)
  • regional variation in white matter
    • volumetric studies
    • tractography
  • during youth, cortical white matter is extensively remodeled and expands a lot
  • evidence from DTI suggests asyncrhonous maturational timing
  • limitation: anatomically pre-defined atlases
    • discrete anatomical bundles = discrete anatomical regions
  • what’s missing: bundle-based approaches do not caputure spatial covariance that may exist across distribute white matter locations
  • certain areas may share similar strutural profiles even though they belong to diff. anatomical bundles
  • structure of subregiosn wtihin the same bundle may vary significantly.
  • spatial organization is likely driven by syndrhonized matural processes
  • dataset: philadelphia neurodevelopmental cohort (PNC)
    • white matter structure has many crossing fibers that aren’t well represented in conventional methods. use fixel-based analysis, more accurate than conventional DTI
    • fixels (FBA) - higher resolution, increase feature space - allows for more accurate white matter structure
  • how do you capture covariance across white matter structure?
  • non-negative matrix factorizatoin (NMF): similar to PCA, but doesn’t maximize variation in first few components, but get a parts-based representation.
  • delineate associations with age and cognition
    • use generalized additive models
  • results:
    • delineate the spatial and temporal organization
    • 14 components: capture white matter structure
  • white matter structure components show widespread development: corpus callosum, arcuate, and splenium
  • developmental gradient: early maturation, inferior and posterior
  • superior cerebellum, vermis, cingulum, early maturation of white matter
  • intermediate regions: pareitnal, splenium (12-16 years maturation)
  • late maturation: end of adolescence (anterior and superior matures later)
  • there might be a hierarchical pattern of age-related changes in white matter structure

Session 2: Beyond the Neocortex

  • Striatal gradients in health and disease
  • non-invasive method for investigating the dopaminergic system (in striatum)
  • biomarker for altered dopaminergic signaling at the single-subject level
  • psychosis: PET studies indicate increased striatal dopaminergic signaling in psychosis.
  • rs-fMRI studies have linked altered cortico-striatal connectivity to both psychosis and subclinical psychosis-like experiences
  • analysis of both 2nd cortico-striatal and 3rd dopaminergic connectivity gradient
  • psychosis 2nd connectivity gradient
  • next steps: investigate striatal gradients in other dopamine-associated conditions such as ASD and ADHD.
    • push connectompic mapping in the temporal domain
  • looking at other regions, find the same pattern - makes sense that you would find this

CMI Autism Spectrum Center Director: Adriana Di Martino - https://childmind.org/science/fundamental-neuroscience/autism-center/ Breanne Kearney bkearne3@uwo.ca - send paper on fear conditioning in the cerebellum

Session3 3: Flash Talks

  • identifying spatiotemporal changes in cortical neurodevelopment in postmortem and in vivo data
  • multi-modal, area- and depth-wise characterization of cortical structure throughout development
  • baby connectome project
  • generation of a normative model of intracortical development
  • microstructure increases across cortical depths, but increases are selectively more pronounced in deeper layers

hierarchical neurodevelopment * emergence of sensory-association axis

Session 4: AI

  • Sofie valk group

  • excitation and inhibition

  • brain changes in adolescence: changes in excitatory and inhibitory balance - implication for neurodevelopmental disorders

  • lots of E/I balance comes from animal models

  • challenges of studying excitatoin and inhibition in humans

  • simulations that map on to imaging data

  • using biophysical network modeling of rs-fMRI, how does excitation-inhibition develop?

  • datasets: PNC (n=764) and IMAGEN (n=149, longitudinal)

  • model-simulation optimization: how do we go from imaging data to E/I balance in each area?

  • development: sensorimotor areas develop early and association matures later

Sensorimotor-association axis and sex differences

  • Sofie valk group
  • developmental and evolutionary expansion
  • inter-individual variability of S-A axis is lacking
    • focus on sex differences
    • intra-individual differences
  • although brain size shows robust sex differences, it is unclear whether it differentially shapes functioanl cortical organization between sexes through altered cortical mophometry
  • brain size is a major scaling factor across evolution and development
  • sex differences present at the poles of the S-A axis. females > DMN and males > sensorimotor network
  • differences in cortical mophometry - responsible for differences in cortical organization?
  • cortical mophometry supports function
    • primary regions - short range connections
    • association - long range connections
  • sex differences in S-A axis?
    • morphometric correlates of S-A axis
    • test first for effect of brain size on S-A axis (ICV, TBC, Total SA) - total SA most widespread effect on S-A axis
    • sex differences in 23% of regions - DMN, frontoparietal
  • sex differences in S-A reflect differences in morphometric correlates?
    • sex differences in cortical morphometry explain little of the variance in S-A axis
  • S-A axis of multimodal organizational hierachy
  • sex differences in 1.2% of all functional connectiviyt measures (schaefer x schaefer networks)
  • females show a more integrated DMN compared to males
  • sex dfiferences in S-A axis do not apear to be systematically explained by diffferneces in brain size, microstructural organization or mean geodesic distance
  • intra-individual differences:
    • Dense sampling - MyConnectome study and Midnight scan club
  • mechanisms underpinning functional variability?
    • the brain interacts with the endocrine system
    • pritschet et al. (2020) and Grotzinger et al. (2023) - jacobs lab data
    • intra-/inter-individual differences in variability
    • inter-individaul differences in variability - higher variability in male subject (case study)
    • local-level effects on S-A axis loadings
  • don’t overinterpret results - there are more similarities than differences. the brain is not sexually dimorphic.

Using a neural state-space to understand cognition and behavior

  • Samyogita Hardikar
  • Gradients allow us to summarize variability in a compact manner
  • Margulies et al. (2016) PNAS - first gradient paper. Axes of functional variation in the brain. Motor-to-association
  • Third gradient separates systems DMN from multiple-demand networks
  • Axes of functional variation in the brain
  • brain-wide associations of personality traits only possible with lots of data
  • do mass-univariate approaches give a “brain-wide” picture
    • no, just smootheed group regions
  • can rest/any one task condition provide all the information?
  • personality dimensions can be conceptualized as “if-then” rules
    • not all traits will be expressed in every given situation
    • intrinsic connectivity (functional organization) + context
  • HCP Young-Adult s1200
    • task fMRI: WM, Gambling, Motor, Language, social cognition, relational processing, emotion
    • trait measures (neo-ffi) - big 5 traits: neuroticism, openness, conscientiousness, extraversion, agreeableness (minimally preprocessed 2mm smoothing CIFTI)
    • neural state-space: primary-associatoin; motor-visual; DMN-FPN

Procrustes alignment

  • solve the issue of gradient alignment