\( \def\RR{\bf R} \def\real{\mathbb{R}} \def\bold#1{\bf #1} \def\d{\mbox{Cord}} \def\hd{\widehat \mbox{Cord}} \DeclareMathOperator{\cov}{cov} \DeclareMathOperator{\var}{var} \DeclareMathOperator{\cor}{cor} \newcommand{\ac}[1]{\left\{#1\right\}} \)

Estimating Brain Pathways Using Large-scale Multilevel Models

Xi (Rossi) LUO

Brown University
Department of Biostatistics
Center for Statistical Sciences
Computation in Brain and Mind
Brown Institute for Brain Science
Brown Data Science Initiative
ABCD Research Group

ICSA, Chicago, USA
June, 2017

Funding: NIH R01EB022911, P20GM103645, P01AA019072; NSF/DMS (Big Data) 1557467


Yi Zhao
(Brown PhD 2017, starting postdoc at JHU Biostat)
This work received the 2015 ENAR distinguished paper award.

Brown University

Math Problem:
Infer Causality from Observational Data Only

Many unmodeled factors: vascular, respiratory, mind wandering, ...

Stop/Go Task fMRI

  • Task fMRI: performs tasks under brain scanning
  • Randomized stop/go task:
    • press button if "go";
    • withhold pressing if "stop"
  • Not resting-state: "rest" in scanner

Goal: causal mechanisms from stimuli to motor cortex

fMRI data: blood-oxygen-level dependent (BOLD) signals from each cube/voxel (~millimeters), $10^5$ ~ $10^6$ voxels in total.

Raw Data: Motor Region

Time (seconds)

Black: fMRI BOLD activity
Blue: stop onset times; Red: go onset times

fMRI Studies

Sub 1, Sess 1

Time 1



Sub i, Sess j

Sub ~100, Sess ~4

Our data: 96 subjects × 4 sessions × 100 trials × 2 regions

Question: can " big and complex" fMRI reveal causality?

Brain Networks

Functional/Effective Connectivity

nodes/connections removed to enhance visualization

Network Model with Stimulus

Goal: quantify effects stimulipreSMAPMC regions Duann, Ide, Luo, Li (2009). J of Neurosci


Mediation Analysis and SEM

$$\begin{align*}M &= Z a + \overbrace{U + \epsilon_1}^{E_1}\\ R &= Z c + M b + \underbrace{U g + \epsilon_2}_{E_2}, \quad \epsilon_1 \bot \epsilon_2\end{align*}$$
  • Indirect effect: $a \times b$; Direct effect: $c$
  • Correlated errors: $\delta = \cor(E_1, E_2) \ne 0$ if $U\ne 0$

Mediation Analysis in fMRI

  • Mediation analysis (usually assuming $U=0$)Baron&Kenny, 86; Sobel, 82; Holland 88; Preacher&Hayes 08; Imai et al, 10; VanderWeele, 15;...
  • Parametric Wager et al, 09 and functional Lindquist, 12 mediation, under (approx.) independent errors
    • Stimulus $\rightarrow$ brain $\rightarrow$ user reported ratings, one brain mediator
    • Assuming $U=0$ between ratings and brain
  • Multiple mediator and multiple pathways
    • Dimension reduction by arXiv1511.09354Chen, Crainiceanu, Ogburn, Caffo, Wager, Lindquist, 15
    • Pathway Lasso penalization Zhao, Luo, 16 (Session #159, Wed 11:45am)
  • This talk: causal estimation under $U\ne 0$ (its effect size $\delta \ne 0$) when modeling two brain regions


  • Activation: stimulus $\rightarrow$ brain region activity
  • Connectivity: one brain region $\rightarrow$ another region
    • Whether not two or more brain regions "correlate"
  • Pathway: stimulus $\rightarrow$ brain region A $\rightarrow$ region B
  • Strong path: strong activation and strong conn
  • Zero path: zero activation or zero conn, including
    • Zero activation + strong conn = zero
    • Strong activation + zero conn = zero

Existing Approaches for $\delta \ne 0$

  • Assuming $\delta=0$
    • Assumption "too strong" for most cases Imai et al, 10
  • Sensitivity plot: "guessing" $\delta$
  • Simplify models: e.g. $c=0$ via instrumental variable
  • Adjust (e.g. motion) if possible Sobel, Lindquist, 14
  • Use Bayesian prior or regularization

Positive or negative effects depending on subjective choice of $\delta$


  • Recall 3-level data: subjects, sessions, trials
  • Correlated mediation model for trials, mixed model for mediation effects among higher levels
    • Special case: 2-level data, mixed becomes anova
  • We will optimize the multilevel likelihood $$\underbrace{\sum_{\mbox{Sub }i} \sum_{\mbox{Sess }k}\ell(\mbox{brain activities in trials} | A_{ik}, B_{ik}, C_{ik}, \Theta_{ik})}_{\mbox{first level likelihood}} \\ + \underbrace{\sum_i \ell(A_{i1},\dotsc,A_{iK}, B_{i1},\dotsc | A, B, C, \Theta)}_{\mbox{second and third level likelihood}}$$


  • Unmeasured confounding and causal inference
    • Usually impossible in many other statistical models
  • Prove our model is identifiable or our multilevel likelihood has a unique maximum
    • Usually the likelihood is mutlimodal in many other cases
  • Prove our MLE is unbiased and consistent under minimal assumptions

Our Level-1 Model

  • Bivariate single-level model $$\begin{align*}\begin{pmatrix}M & R\end{pmatrix} & = \begin{pmatrix}Z & M\end{pmatrix} \begin{pmatrix}a & c\\ 0 & b \end{pmatrix} + \begin{pmatrix}E_{1} & E_{2}\end{pmatrix}\\ (E_1, E_2) & \sim MVN(\boldsymbol{0}, \boldsymbol{\Sigma}) \quad \boldsymbol{\Sigma}=\begin{pmatrix}\sigma_{1}^{2} & \delta\sigma_{1}\sigma_{2}\\ \delta\sigma_{1}\sigma_{2} & \sigma_{2}^{2} \end{pmatrix} \end{align*}$$
  • Estimate $(a,b,c,\boldsymbol{\Sigma})$ via ML (a lot of handwaving)
  • We solve an optimization with constraints
    • Different than running two regressions

Causal Interpretation

  • Prove causal using potential outcomes Neyman, 23; Rubin, 74
  • Causal inference intuition $$Z \rightarrow \begin{pmatrix} M \\ R \end{pmatrix}$$
  • Others assume (e.g. randomized brain stimulation) $$r_{i}\left(z_{i}^{\prime},m_{i}\right)\bot m_{i}\left(z_{i}\right)|Z_{i}=z_{i}$$
  • We do not need this assumption when we can estimate $\delta$


  • Theorem: Given $\delta$, unique maximizer of likelihood, expressed in closed form
  • Theorem: Given $\delta$, our estimator is root-n consistent and efficient
  • Bias (and variance) depends on $\delta$

"Tragedy" of ML

Theorem: The maximum profile likelihood value is the same for every $\delta \in (-1,+1)$.
  • Likelihood provides zero info about $\delta$
  • Cannot simply apply priors on $\delta$

Two different models generate the same single-trial BOLD activations if only observing $Z$, $M$, and $R$

without measuring $U$

Our Higher Level Models

  • Cannot identify $\delta$ from single sub and single sess (see our theorem)
  • Intuition: leverage complex data structure to infer $\delta$

Some Details

  • Level-1 model for each sub and each sess $$\begin{pmatrix}{M}_{ik} & {R}_{ik}\end{pmatrix}=\begin{pmatrix}{Z}_{ik} & {M}_{ik}\end{pmatrix}\begin{pmatrix}{a}_{ik} & {b}_{ik}\\ 0 & {c}_{ik} \end{pmatrix}+\begin{pmatrix}{E}_{1_{ik}} & {E}_{2_{ik}}\end{pmatrix}$$
  • Limited variability in $\delta$ across sub/sess
  • Random effect model cf AFNI, FSL, SPM, and etc $$\begin{pmatrix}{A}_{ik}\\ {B}_{ik}\\ {C}_{ik} \end{pmatrix}=\begin{pmatrix}{A}\\ {B}\\ {C} \end{pmatrix}+\begin{pmatrix}\alpha_{i}\\ \beta_{i}\\ \gamma_{i} \end{pmatrix}+\begin{pmatrix}\epsilon_{ik}^{{A}}\\ \epsilon_{ik}^{{B}}\\ \epsilon_{ik}^{{C}} \end{pmatrix}=b+u_{i}+\eta_{ik}$$

Algorithm 1: Two-stage Fitting

  • Stage 1: fit $(\hat{A}_{ik}(\delta), \hat{B}_{ik}(\delta), \hat{C}_{ik}(\delta))$ for each $i$ and $k$ for varying $\delta$ using our step 1 single-level model
  • Stage 2: Find $\hat{\delta}$ that $(\hat{A}_{ik}(\hat{\delta}), \hat{B}_{ik}(\hat{\delta}), \hat{C}_{ik}(\hat{\delta}))$ yields maximum likelihood for random effects model
  • Small-scale computing

Theory for SEM and Confounding

Theorem: Under certain regularity conditions, asymptotically, the joint multilevel likelihood has a unique maximum and the maximizer is consistent for $\delta$.

Contributions: data-driven estimation of confounding and consistency proof for SEM

Alternative Likelihood under Our Framework

  • Optimize all parameters in joint likelihood $$\begin{align*} &\sum_{i=1}^{N}\sum_{k=1}^{K}\log\Pr\left(R_{ik},M_{ik}|Z_{ik},\delta,b_{ik},\sigma_{1_{ik}},\sigma_{2_{ik}}\right)\quad \mbox{Data}\\ & + \sum_{i=1}^{N}\sum_{k=1}^{K}\log\Pr\left(b_{ik}|u_{i},b,\boldsymbol{\Lambda}\right)\quad \mbox{Subject variation}\\ & +\sum_{i=1}^{N}\log\Pr\left(u_{i}|\boldsymbol{{\Psi}}\right) \quad \mbox{Prior}\end{align*}$$
  • Large computation: $5NK + 3N + 11 > 2000$ paras

Algorithm 2

Theorem: The joint likelihood is conditional convex for groups of parameters.
Algorithm: block coordinate descent with projections.

Leverage conditional convexity to reduce computation

details in our paper


Method Comparison

  • Single level models
    • BKBaron, Kenny, 86: assuming $\delta = 0$
    • CMA-$\delta$ (our single-level method): true $\delta$ (oracle)
  • Mixed effects multilevel models
    • KKBKenny et al, 03: assuming all $\delta = 0$
    • CMA-ts (ours): estimating $\delta$, two-stage fitting
    • CMA-h, CMA-m (ours): estimating $\delta$, large-scale fitting
  • Simulate data with varying $\delta$

Low bias for $\delta$

Low bias for effects

Our large-scale multilevel model (CMA) performs best;
Bias can be large (100%) without estimating $\delta$

Real Data


  • Random stimuli and scalar activations for each trial
  • Stimuli: go = press; stop = not pressing
    • stop is expected to suppress motor area (PMC)
  • 96 subjects × 4 sessions × 100 trials × 2 regions
  • Temporal dependence removed using whitened data and single-trial beta deconvolutionWager et al, 08

Identifiability and Bias

Unique $\delta$ for ML

Estimates doubled

  • Ours show stop directly suppresses PMC significantly while other methods fail to recover this
  • Motion (part of $U$) correction decreases our $\delta$ estimate but do not change our causal estimates


  • Leverage complex data for causal fMRI modeling
  • Approach:
    • large-scale causal SEM
    • big multilevel data
    • machine learning/optimization
  • Theory: identifiability, convergence, and computation
  • Result: reduced bias and causal interpretation


  • Extensions: functional models, multiple pathwaysPathway Lasso, our arXiv 1603.07749, temporal 2017 JSM Mental Health Student Paper Award Section
  • Manuscript: arXiv 1410.7217
  • R package: macc


Thank you!

Comments? Questions?

Slides at: BrainDataScience.com

or BigComplexData.com