arxiv: 2605.14056 · v1 · submitted 2026-05-13 · 📊 stat.ME · stat.AP

Recognition: 2 theorem links

· Lean Theorem

An MCMC-Based Method for Dynamic Causal Modeling of Effective Connectivity in Functional MRI

Kaitlyn R. Fales , Hyebin Song , Nicole A. Lazar

Authors on Pith no claims yet

Pith reviewed 2026-05-15 02:07 UTC · model grok-4.3

classification 📊 stat.ME stat.AP

keywords dynamic causal modelingeffective connectivityfMRIMCMCNo-U-Turn Sampleruncertainty quantificationidentifiabilityBOLD response

0 comments

The pith

CDCM uses MCMC and a simpler observation model to estimate fMRI effective connectivity with consistent parameters and reliable uncertainty.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Canonical DCM (CDCM), an MCMC-based alternative to standard dynamic causal modeling for directional brain connectivity in fMRI. It replaces the usual complex neural-hemodynamic observation model with a simpler one that permits a piecewise analytic solution to the underlying neural ODE. This change supports faster computation, explicit conditions for parameter identifiability, and posterior sampling via the No-U-Turn Sampler. Tests on simulated data and public fMRI datasets from the Human Connectome Project show consistent recovery of input-related parameters and well-calibrated uncertainty. A reader would care because the approach directly tackles known practical problems in conventional DCM, such as inexact solutions and underestimated uncertainty, while remaining grounded in the same state-space framework.

Core claim

CDCM is a Markov chain Monte Carlo method for dynamic causal modeling that adopts a simpler observation model for the BOLD signal and employs the No-U-Turn Sampler for posterior inference. The simpler model admits a piecewise analytic solution to the neural ordinary differential equation, which raises computational efficiency and permits explicit derivation of sufficient conditions for parameter identifiability. On both simulated and real fMRI data the method yields reliable uncertainty quantification together with consistent estimation of parameters that govern experimental inputs.

What carries the argument

Canonical DCM (CDCM), which pairs a reduced observation model with No-U-Turn Sampler MCMC to solve the neural state-space equations for effective connectivity.

If this is right

CDCM supplies better-calibrated posterior uncertainty than variational Bayes approaches for the same state-space model.
The piecewise analytic solution reduces the need for numerical integration at each MCMC step.
Explicit identifiability conditions derived from the simplified model can be checked before fitting real data.
Connectivity estimates remain replicable across small- and large-scale public neuroimaging datasets.
Parameters tied to experimental inputs are recovered consistently between simulated and empirical runs.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same reduced observation model could be paired with other sampling schemes or optimization routines beyond No-U-Turn MCMC.
Replicability results suggest CDCM might serve as a standardized baseline for comparing effective-connectivity findings across laboratories.
If the simpler model proves adequate, extensions could incorporate additional covariates or multi-subject hierarchical structures without further loss of analytic tractability.
The identifiability conditions may guide experimental design by indicating which input contrasts are most informative for a given network.

Load-bearing premise

The simpler observation model captures the essential neural-hemodynamic dynamics without introducing bias into the connectivity estimates.

What would settle it

A controlled simulation in which connectivity parameters recovered by CDCM systematically deviate from those recovered by the full neural-hemodynamic model while ground-truth values are known.

Figures

Figures reproduced from arXiv: 2605.14056 by Hyebin Song, Kaitlyn R. Fales, Nicole A. Lazar.

**Figure 1.** Figure 1: Hillebrandt et al 51 group-level DCM using Q2 HCP data, with each parameter shown being the grand mean estimate across all subjects, sessions, and hemispheres. The grand means are reported as no session (RL, LR phase encoding) or hemisphere (right, left) effects are found from conducting repeated-measures ANOVAs on pairs of connections. Some connections shown in the state-space hypothesis do not have mean … view at source ↗

**Figure 2.** Figure 2: Schematic representation of Proposition 1. Constant stimulus blocks induce affine dynamics, and the trajectory z(t) is constructed by solving each block and propagating via z(t (b) ). 3.3 Individual and Group-Level Parametric Estimation Procedure Through the simplification of the observation model, CDCM involves fewer parameters and a simpler structure, enabling an exact parametric estimation procedure tha… view at source ↗

**Figure 3.** Figure 3: A network representation of the model hypothesis used to conduct the simulations [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗

**Figure 4.** Figure 4: (Top) Average posterior mean estimates and 95% HPD intervals across 50 simu [PITH_FULL_IMAGE:figures/full_fig_p023_4.png] view at source ↗

**Figure 5.** Figure 5: A network representation of a DCM hypothesis from the attention to visual motion [PITH_FULL_IMAGE:figures/full_fig_p025_5.png] view at source ↗

**Figure 6.** Figure 6: Attention to visual motion posterior mean estimates and 95% HPD intervals for [PITH_FULL_IMAGE:figures/full_fig_p026_6.png] view at source ↗

**Figure 7.** Figure 7: The estimated versus observed BOLD signal in each ROI between CDCM (blue) [PITH_FULL_IMAGE:figures/full_fig_p028_7.png] view at source ↗

**Figure 8.** Figure 8: Group-level posterior mean estimates and 95% HPD interval comparison between [PITH_FULL_IMAGE:figures/full_fig_p031_8.png] view at source ↗

**Figure 9.** Figure 9: Resulting average group-level models across sessions and hemispheres using CDCM [PITH_FULL_IMAGE:figures/full_fig_p032_9.png] view at source ↗

read the original abstract

Effective connectivity analysis in functional magnetic resonance imaging (fMRI) studies directional interactions among brain regions and experimental stimuli. Dynamic causal modeling (DCM) is a widely used method to estimate effective connectivity, based on a state-space representation consisting of a latent neural signal model and an observation model transforming the neural signal into the observed blood-oxygen-level-dependent (BOLD) response. A standard DCM combines ordinary differential equation (ODE) dynamics for the latent signal with a complex neural-hemodynamic system for the observation model, and typically uses variational Bayes for parameter estimation. While physically well-motivated, this approach can lead to practical challenges such as inexact solutions and underestimated uncertainty. We introduce Canonical DCM (CDCM), a Markov chain Monte Carlo (MCMC)-based method that adopts a simpler observation model and the No-U-Turn Sampler for posterior sampling. The simpler observation model admits a piecewise analytic solution to the neural ODE, increasing computational efficiency and enabling explicit derivation of sufficient conditions for parameter identifiability. The results indicate that CDCM provides reliable uncertainty quantification and consistent estimation of parameters related to experimental inputs for simulated and real data. We use publicly available data from the Wellcome Centre for Human Neuroimaging and the Human Connectome Project (HCP) to benchmark CDCM against standard DCM methods and examine replicability of estimated connectivity patterns in small- and large-scale neuroimaging settings.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

CDCM swaps variational Bayes for NUTS MCMC plus a simpler observation model that admits piecewise analytic ODE solutions, and the checks on simulated and public data look internally consistent.

read the letter

The paper's core move is replacing the standard variational Bayes pipeline in DCM with No-U-Turn Sampler MCMC and a reduced observation model whose neural ODE can be solved analytically in pieces. This change is meant to deliver better-calibrated uncertainty on the connectivity parameters while keeping the linear input effects intact. They derive explicit identifiability conditions for the reduced model and show that it preserves the quantities they care about for experimental inputs. On simulated data they report posterior coverage checks against ground truth, and on the Wellcome and HCP datasets they compare point estimates and replicability patterns against conventional DCM. Those steps are straightforward and use public data, which is helpful. The main soft spot is whether the stripped-down observation model stays adequate once the hemodynamics get more complicated than the linear-input regime they target; the paper argues it does, but the claim rests on how closely the posteriors match standard DCM outputs in the real-data benchmarks. If the MCMC diagnostics and coverage numbers hold up in the full supplement, the uncertainty-quantification improvement is real. This is useful for people already working on effective connectivity who want an MCMC option or tighter uncertainty intervals. It is worth sending to peer review because the method is self-contained, the evaluation protocol is reproducible, and the central consistency claim can be checked directly against the reported simulations and public datasets.

Referee Report

1 major / 2 minor

Summary. The manuscript introduces Canonical DCM (CDCM), an MCMC-based alternative to standard dynamic causal modeling for effective connectivity in fMRI. It replaces the conventional complex neural-hemodynamic observation model with a simpler one that admits a piecewise analytic solution to the neural ODE, employs the No-U-Turn Sampler (NUTS) for posterior sampling, derives explicit identifiability conditions, and reports reliable uncertainty quantification together with consistent estimation of experimental-input parameters on both simulated data and public real datasets from the Wellcome Centre and HCP.

Significance. If the central claims hold, CDCM would supply a computationally efficient route to better-calibrated posterior uncertainty in effective-connectivity estimates, addressing documented limitations of variational Bayes in standard DCM. The explicit identifiability analysis and use of publicly available benchmarking data constitute concrete strengths that could improve replicability assessments in small- and large-scale neuroimaging studies.

major comments (1)

[§4.3] §4.3 (simulation protocol): the reported posterior coverage for input parameters is shown only for the reduced observation model; a direct comparison of bias in connectivity estimates between the full hemodynamic model and the simplified model is needed to substantiate the claim that the simplification introduces no material bias in the quantities the method targets.

minor comments (2)

[Eq. 8–10] Notation for the piecewise solution (Eq. 8–10) should explicitly state the switching times and continuity conditions to avoid ambiguity when readers implement the analytic form.
[§5] The HCP and Wellcome benchmarking sections would benefit from a table summarizing effective sample sizes and Gelman–Rubin statistics across all reported runs.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We are grateful to the referee for their positive evaluation of our manuscript and for the constructive feedback. We believe the suggested addition will further strengthen the presentation of our results.

read point-by-point responses

Referee: [§4.3] §4.3 (simulation protocol): the reported posterior coverage for input parameters is shown only for the reduced observation model; a direct comparison of bias in connectivity estimates between the full hemodynamic model and the simplified model is needed to substantiate the claim that the simplification introduces no material bias in the quantities the method targets.

Authors: We thank the referee for this observation. While our simulations demonstrate consistent estimation and good coverage for the input parameters under the simplified model, we acknowledge that a head-to-head comparison with the full model would provide stronger evidence against material bias in connectivity estimates. In the revised version, we will extend §4.3 to include such a comparison. Specifically, we will simulate data under the full hemodynamic model and then estimate parameters using both the full model (via standard DCM) and our CDCM simplified model, reporting bias and coverage for the connectivity parameters. This will allow readers to assess any differences introduced by the simplification. We note that the full model estimation will be performed using the available DCM software for fairness. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper introduces an independent MCMC procedure (NUTS sampler) and a deliberately simplified observation model whose piecewise-analytic solution is derived directly from the reduced ODE rather than fitted to prior outputs. Identifiability conditions are obtained explicitly from the analytic form, simulation protocols use ground-truth parameters independent of the fitted model, and benchmarking on HCP/Wellcome data compares against external standard DCM implementations. No step reduces a claimed prediction to a self-defined quantity, a fitted input renamed as prediction, or a load-bearing self-citation chain; the derivation remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The approach rests on standard DCM state-space assumptions plus the new claim that the simplified observation model permits an exact piecewise solution to the neural ODE.

free parameters (1)

DCM connectivity and input parameters
Estimated from data via MCMC; exact count and priors not specified in abstract.

axioms (1)

domain assumption Neural dynamics follow ODEs that admit piecewise analytic solution under the simplified observation model
Invoked to justify computational efficiency and identifiability conditions.

pith-pipeline@v0.9.0 · 5552 in / 1134 out tokens · 49591 ms · 2026-05-15T02:07:31.057436+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

81 extracted references · 81 canonical work pages

[1]

Identification of affine dynamical systems from a single trajectory.Inverse Problems

Duan X, Rubin J, Swigon D. Identification of affine dynamical systems from a single trajectory.Inverse Problems. 2020;36:no. 085004

work page 2020
[2]

Identifiability of linear and linear-in-parameters dy- namical systems from a single trajectory.SIAM Journal on Applied Dynamical Systems

Stanhope S, Rubin J, Swigon D. Identifiability of linear and linear-in-parameters dy- namical systems from a single trajectory.SIAM Journal on Applied Dynamical Systems. 2014;13:1792-1815

work page 2014
[3]

Perspectives on system identification.Annual Reviews in Control

Ljung L. Perspectives on system identification.Annual Reviews in Control. 2010;34:1-12

work page 2010
[4]

System identification of nonlinear state-space models

Schön T, Wills A, Ninness B. System identification of nonlinear state-space models. Automatica. 2011;47:39-49

work page 2011
[5]

Identifiability and asymp- totics in learning homogeneous linear ODE systems from discrete observations.Journal of Machine Learning Research

Wang Y, Huang W, Gong M, Geng X, Liu T, Zhang K, et al. Identifiability and asymp- totics in learning homogeneous linear ODE systems from discrete observations.Journal of Machine Learning Research. 2024;25:1-50

work page 2024
[6]

On identifiability of nonlinear ODE models and applications in viral dynamics.SIAM Review

Miao H, Xia X, Perelson A, Wu H. On identifiability of nonlinear ODE models and applications in viral dynamics.SIAM Review. 2011;53:3-39

work page 2011
[7]

Observability and structural identifiability of nonlinear biological systems

Villaverde A. Observability and structural identifiability of nonlinear biological systems. Complexity. 2019;2019:no. 8497093

work page 2019
[8]

On the general theory of control systems.IFAC Proceedings Volumes

Kalman R. On the general theory of control systems.IFAC Proceedings Volumes. 1960;1:491-502. 35

work page 1960
[9]

Nonlinear controllability and observability.IEEE Transactions on Automatic Control

Hermann R, Krener A. Nonlinear controllability and observability.IEEE Transactions on Automatic Control. 1977;22:728-740

work page 1977
[10]

On structural identifiability.Mathematical Biosciences

Bellman R, Åström K. On structural identifiability.Mathematical Biosciences. 1970;7:329-339

work page 1970
[11]

Structural and practical identifiability analysis of partially observed dynamical models.Bioinfor- matics

Raue A, Kreutz C, Maiwald T, Bachmann J, Schilling M, Klingmuller U, et al. Structural and practical identifiability analysis of partially observed dynamical models.Bioinfor- matics. 2009;25:1923-1929

work page 2009
[12]

Englewood Cliffs, NJ: Prentice Hall; 1980

Kailath T.Linear Systems. Englewood Cliffs, NJ: Prentice Hall; 1980

work page 1980
[13]

Structural identifiability of systems biology models: A critical comparison of methods.PLOS ONE

Chis OT, Banga J, Balsa-Canto E. Structural identifiability of systems biology models: A critical comparison of methods.PLOS ONE. 2011;6:no. e27755

work page 2011
[14]

On structural and practical identifia- bility.Current Opinion in Systems Biology

Wieland F, Hauber F, Rosenblatt M, Timmer J. On structural and practical identifia- bility.Current Opinion in Systems Biology. 2021;25:60-69

work page 2021
[15]

Computing all identifiable functions of parameters for ODE models.Systems & Control Letters

Ovchinnikov A, Pillay A, Pogudin G, Scanlon T. Computing all identifiable functions of parameters for ODE models.Systems & Control Letters. 2021;157:no. 105030

work page 2021
[16]

Identifiable specializations for ODE models.Systems & Control Letters

Ovchinnikov A, Pillay A, Pogudin G, Scanlon T. Identifiable specializations for ODE models.Systems & Control Letters. 2025;204:no. 106226

work page 2025
[17]

Dynamic causal modelling.NeuroImage

Friston K, Harrison L, Penny W. Dynamic causal modelling.NeuroImage. 2003;19:1273- 1302

work page 2003
[18]

A guide to group effective connectivity analysis, part 1: First level analysis with DCM for fMRI.Neu- roImage

Zeidman P, Jafarian A, Corbin N, Seghier M, Razi A, Price C, et al. A guide to group effective connectivity analysis, part 1: First level analysis with DCM for fMRI.Neu- roImage. 2019;200:174-190

work page 2019
[19]

Assessing parameter identifiability for dynamic causal modeling of fMRI data.Frontiers in Neuroscience

Arand C, Scheller E, Seeber B, Timmer J, Klöppel S, Schelter B. Assessing parameter identifiability for dynamic causal modeling of fMRI data.Frontiers in Neuroscience. 2015;9:no. 43

work page 2015
[20]

Ten simple rules for dynamic causal modeling.NeuroImage

Stephan K, Penny W, Moran R, den Ouden H, Daunizeau J, Friston K. Ten simple rules for dynamic causal modeling.NeuroImage. 2010;49:3099-3109

work page 2010
[21]

Understanding DCM: Ten simple rules for the clinician.NeuroIm- age

Kahan J, Foltynie T. Understanding DCM: Ten simple rules for the clinician.NeuroIm- age. 2013;83:542-549

work page 2013
[22]

Dynamic causal modelling for fMRI: A two-state model.NeuroImage

Marreiros A, Kiebel S, Friston K. Dynamic causal modelling for fMRI: A two-state model.NeuroImage. 2008;39:269-278

work page 2008
[23]

Dynamic causal modelling: A critical review of the biophysical and statistical foundations.NeuroImage

Daunizeau J, David O, Stephan K. Dynamic causal modelling: A critical review of the biophysical and statistical foundations.NeuroImage. 2011;58:312-322

work page 2011
[24]

Regression dynamic causal modeling for resting-state fMRI.Human Brain Mapping

Frässle S, Harrison S, Heinzle J, Clementz B, Tamminga C, Sweeney J, et al. Regression dynamic causal modeling for resting-state fMRI.Human Brain Mapping. 2021;42:2159- 2180. 36

work page 2021
[25]

Multiple- shooting adjoint method for whole-brain dynamic causal modeling

Zhuang J, Dvornek N, Tatikonda S, Papademetris X, Ventola P, Duncan J. Multiple- shooting adjoint method for whole-brain dynamic causal modeling. In:Proceedings of the International Conference on Information Processing in Medical Imaging (IPMI). 2021:58–70

work page 2021
[26]

Spectral dynamic causal modeling: A didactic introduction and its relationship with functional connectivity.Network Neuroscience

Novelli L, Friston K, Razi A. Spectral dynamic causal modeling: A didactic introduction and its relationship with functional connectivity.Network Neuroscience. 2024;8:178-202

work page 2024
[27]

CausalMamba: Scalable conditional state space models for neural causal inference.arXiv preprint

Bae S, Cha J. CausalMamba: Scalable conditional state space models for neural causal inference.arXiv preprint. 2025. ArXiv:2510.17318. Available from:https://arxiv. org/abs/2510.17318

work page arXiv 2025
[28]

Nonlinear responses in fMRI: The Balloon model, Volterra kernels, and other hemodynamics.NeuroImage

Friston K, Mechelli A, Turner R, Price C. Nonlinear responses in fMRI: The Balloon model, Volterra kernels, and other hemodynamics.NeuroImage. 2000;12:466-477

work page 2000
[29]

Bayesian estimation of dynamical systems: An application to fMRI.Neu- roImage

Friston K. Bayesian estimation of dynamical systems: An application to fMRI.Neu- roImage. 2002;16:513-530

work page 2002
[30]

Comparing hemodynamic models with DCM.NeuroImage

Stephan K, Weiskopf N, Drysdale P, Robinson P, Friston K. Comparing hemodynamic models with DCM.NeuroImage. 2007;38:387-401

work page 2007
[31]

Nonlinear dynamic causal models for fMRI.NeuroImage

Stephan K, Kasper L, Harrison L, Daunizeau J, den Ouden H, Breakspear M, et al. Nonlinear dynamic causal models for fMRI.NeuroImage. 2008;42:649-662

work page 2008
[32]

Stochastic dynamic causal modelling of fMRI data: Should we care about neural noise?NeuroImage

Daunizeau J, Stephan K, Friston K. Stochastic dynamic causal modelling of fMRI data: Should we care about neural noise?NeuroImage. 2012;62:464-481

work page 2012
[33]

Englewood Cliffs, NJ: Prentice Hall; 2002

Khalil H.Nonlinear Systems. Englewood Cliffs, NJ: Prentice Hall; 2002

work page 2002
[34]

New York, NY: Springer; 1998

Sontag E.Mathematical Control Theory: Deterministic Finite Dimensional Systems. New York, NY: Springer; 1998

work page 1998
[35]

Qualitative theory of compartmental systems.SIAM Review

Jacquez J, Simon C. Qualitative theory of compartmental systems.SIAM Review. 1993;35:43-79

work page 1993
[36]

VBA: A probabilistic treatment of nonlinear models for neurobiological and behavioural data.PLOS Computational Biology

Daunizeau J, Adam V, Rigoux L. VBA: A probabilistic treatment of nonlinear models for neurobiological and behavioural data.PLOS Computational Biology. 2014;10:no. e1003441

work page 2014
[37]

Variational free energy and the Laplace approximation.NeuroImage

Friston K, Mattout J, Trujillo-Barreto N, Ashburner J, Penny W. Variational free energy and the Laplace approximation.NeuroImage. 2006;34:220-234

work page 2006
[38]

A primer on variational Laplace.NeuroImage

Zeidman P, Friston K, Parr T. A primer on variational Laplace.NeuroImage. 2023;279:no. 120310

work page 2023
[39]

Variational inference: A review for statisticians

Blei D, Kucukelbir A, McAuliffe J. Variational inference: A review for statisticians. Journal of the American Statistical Association. 2017;112:859-877

work page 2017
[40]

Gradient-free MCMC methods for dynamic causal modelling.NeuroImage

Sengupta B, Friston K, Penny W. Gradient-free MCMC methods for dynamic causal modelling.NeuroImage. 2015;112:375-381. 37

work page 2015
[41]

Gradient-based MCMC samplers for dynamic causal modelling.NeuroImage

Sengupta B, Friston K, Penny W. Gradient-based MCMC samplers for dynamic causal modelling.NeuroImage. 2016;125:1107-1118

work page 2016
[42]

An introduction to thermodynamic integration and application to dynamic causal models.Cognitive Neurodynamics

Aponte E, Yao Y, Raman S, Frässle S, Heinzle J, Penny W, et al. An introduction to thermodynamic integration and application to dynamic causal models.Cognitive Neurodynamics. 2022;16:1-15

work page 2022
[43]

The WU-Minn Human Connectome Project: An overview.NeuroImage

Van Essen D, Smith S, Barch D, Behrens T, Yacoub E, Ugurbil K, et al. The WU-Minn Human Connectome Project: An overview.NeuroImage. 2013;80:62-79

work page 2013
[44]

The minimal preprocessing pipelines for the Human Connectome Project.NeuroImage

Glasser M, Sotiropoulos S, Wilson J, Coalson T, Fischl B, Andersson J, et al. The minimal preprocessing pipelines for the Human Connectome Project.NeuroImage. 2013;80:105-124

work page 2013
[45]

Human Connectome Project; 2025

WU-Minn Human Connectome Project Consortium.HCP-Young Adult 2025 release. Human Connectome Project; 2025. Data release documentation. ConnectomeDB pow- ered by BALSA. Updated processing of task and resting-state fMRI data. Avail- able from:https://www.humanconnectome.org/study/hcp-young-adult/document/ hcp-young-adult-2025-release

work page 2025
[46]

Movement and mind: A functional imaging study of perception and interpretation of complex intentional movement patterns.NeuroImage

Castelli F, Happe F, Frith U, Frith C. Movement and mind: A functional imaging study of perception and interpretation of complex intentional movement patterns.NeuroImage. 2000;12:314-325

work page 2000
[47]

Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes.Brain

Castelli F, Frith C, Happe F, Frith U. Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes.Brain. 2002;125:1839-1849

work page 2002
[48]

Understanding animate agents: Distinct roles for the social network and mirror system.Psychological Science

Wheatley T, Milleville S, Martin A. Understanding animate agents: Distinct roles for the social network and mirror system.Psychological Science. 2007;18:469-474

work page 2007
[49]

Developing the Frith-Happe animations: A quick and objective test of theory of mind for adults with autism.Autism Research

White S, Coniston D, Rogers R, Frith U. Developing the Frith-Happe animations: A quick and objective test of theory of mind for adults with autism.Autism Research. 2011;4:149-154

work page 2011
[50]

Function in the human connectome: Task-fMRI and individual differences in behavior.NeuroImage

Barch D, Burgess G, Harms M, Petersen S, Schlaggar B, Corbetta M, et al. Function in the human connectome: Task-fMRI and individual differences in behavior.NeuroImage. 2013;80:169-189

work page 2013
[51]

Effective connectivity during animacy perception–dynamic causal modelling of Human Connectome Project data.Scientific Reports

Hillebrandt H, Friston K, Blakemore S. Effective connectivity during animacy perception–dynamic causal modelling of Human Connectome Project data.Scientific Reports. 2014;4:no. 6240

work page 2014
[52]

Dynamics of blood flow and oxygenation changes during brain activation: The Balloon model.Magnetic Resonance in Medicine

Buxton R, Wong E, Frank L. Dynamics of blood flow and oxygenation changes during brain activation: The Balloon model.Magnetic Resonance in Medicine. 1998;39:855-864

work page 1998
[53]

San Diego, CA: Academic Press; 2006

Henson R, Friston K.Statistical Parametric Mapping: The Analysis of Functional Brain Images. San Diego, CA: Academic Press; 2006. 38

work page 2006
[54]

Modeling the hemodynamic response function in fMRI: Efficiency, bias and mid-modeling.NeuroImage

Lindquist M, Loh JM, Atlas L, Wager T. Modeling the hemodynamic response function in fMRI: Efficiency, bias and mid-modeling.NeuroImage. 2009;1 Suppl:S187-S198

work page 2009
[55]

Raue A, Kreutz C, Theis F, Timmer J. Joining forces of Bayesian and frequentist methodology: A study for inference in the presence of non-identifiability.Philosophical Transactions Series A, Mathematical, Physical, and Engineering Sciences. 2012;371:no. 20110544

work page 2012
[56]

Comparison of blocked and event-related fMRI designs for pre-surgical language mapping.NeuroImage

Tie Y, Suarez R, Whalen S, Radmanesh A, Norton I, Golby A. Comparison of blocked and event-related fMRI designs for pre-surgical language mapping.NeuroImage. 2009;47 Suppl 2:T107-T115

work page 2009
[57]

The No-U-Turn Sampler: Adaptively setting path lengths in Hamiltonian Monte Carlo.Journal of Machine Learning Research

Hoffman M, Gelman A. The No-U-Turn Sampler: Adaptively setting path lengths in Hamiltonian Monte Carlo.Journal of Machine Learning Research. 2014;15:1593-1623

work page 2014
[58]

Stan; 2024

Stan Development Team.Stan modeling language users guide and reference manual. Stan; 2024. Version 2.35. Available from:https://mc-stan.org

work page 2024
[59]

CRAN; 2024

Gabry J, Češnovar R, Johnson A.cmdstanr: R interface to CmdStan. CRAN; 2024. R package version 0.7.1. Available from:https://mc-stan.org/cmdstanr/

work page 2024
[60]

Pathfinder: Parallel quasi-Newton varia- tional inference.Journal of Machine Learning Research

Zhang L, Carpenter B, Gelman A, Vehtari A. Pathfinder: Parallel quasi-Newton varia- tional inference.Journal of Machine Learning Research. 2022;23:1-49

work page 2022
[61]

Multivariate output Aaalysis for Markov chain Monte Carlo

Vats D, Flegal J, Jones G. Multivariate output Aaalysis for Markov chain Monte Carlo. Biometrika. 2019;106:321-337

work page 2019
[62]

Flegal J, Vats D, Jones G.mcmcse: Monte Carlo standard errors for MCMC. CRAN

work page
[63]

Available from:https://cran.r-project.org/ package=mcmcse

R package version 1.5-1. Available from:https://cran.r-project.org/ package=mcmcse

work page
[64]

Multivariate moment least-squares variance estimators for reversible Markov chains.Journal of Computational and Graphical Statistics

Song H, Berg S. Multivariate moment least-squares variance estimators for reversible Markov chains.Journal of Computational and Graphical Statistics. 2025;34:409-421

work page 2025
[65]

GitHub; 2025

Song H, Berg S.momentLS: Moment-based long-run variance estimation for Markov chains. GitHub; 2025. R package version 0.1.0. Available from:https://github.com/ hsong1/momentLS

work page 2025
[66]

Gelman A, Carlin J, Stern H, Dunson D, Vehtari A, Rubin D.Bayesian Data Analysis. 3rd ed. Boca Raton, FL: CRC Press; 2013

work page 2013
[67]

Bayesian approaches to random-effects meta-analysis.Statis- tics in Medicine

Smith T, Spiegelhalter D. Bayesian approaches to random-effects meta-analysis.Statis- tics in Medicine. 1995;14:389-401

work page 1995
[68]

Functional connectomics from resting-state fMRI.Trends in Cognitive Sciences

Smith S, Vidaurre D, Beckmann C, Glasser M, Jenkinson M, Miller K, et al. Functional connectomics from resting-state fMRI.Trends in Cognitive Sciences. 2013;17:666-682

work page 2013
[69]

Modulation of connectivity in visual pathways by attention: Corti- cal interactions evaluated with structural equation modelling and fMRI.Cerebral Cortex

Büchel C, Friston K. Modulation of connectivity in visual pathways by attention: Corti- cal interactions evaluated with structural equation modelling and fMRI.Cerebral Cortex. 1997;7:768-778. 39

work page 1997
[70]

Nineteen dubious ways to compute the exponential of a matrix, twenty-five years later.SIAM Review

Moler C, Van Loan C. Nineteen dubious ways to compute the exponential of a matrix, twenty-five years later.SIAM Review. 2003;45:3-49

work page 2003
[71]

Wellcome Centre for Human Neuroimaging, University College London; 2020

FIL Methods Group.SPM12 Manual: Statistical parametric mapping. Wellcome Centre for Human Neuroimaging, University College London; 2020. Chapter 36: Dynamic Causal Modeling for fMRI. Available from:https://www.fil.ion.ucl.ac.uk/spm/ doc/spm12_manual.pdf

work page 2020
[72]

Development of abbre- viated nine-item forms of the Raven’s standard progressive matrices test.Assessment

Bilker W, Hansen J, Brensinger C, Richard J, Gur RE, Gur RC. Development of abbre- viated nine-item forms of the Raven’s standard progressive matrices test.Assessment. 2012;19:354-369

work page 2012
[73]

A DCM for resting state fMRI.NeuroImage

Friston K, Kahan J, Biswal B, Razi A. A DCM for resting state fMRI.NeuroImage. 2014;94:396-407

work page 2014
[74]

Variability and reliability of effective connectivity within the core default mode network: A multi-site longitudinal spectral DCM study.NeuroImage

Almgren H, Van de Steen F, Kühn S, Razi A, Friston K, Marinazzo D. Variability and reliability of effective connectivity within the core default mode network: A multi-site longitudinal spectral DCM study.NeuroImage. 2018;183:757-768

work page 2018
[75]

Eo J, Kang J, Youn T, Park HJ. Neuropharmacological computational analysis of lon- gitudinal electroencephalograms in clozapine-treated patients with schizophrenia using hierarchical dynamic causal modeling.NeuroImage. 2023;275:no. 120161

work page 2023
[76]

Estimating effective connectivity in Alzheimer’s disease progression: A dynamic causal modeling study.Frontiers in Human Neuroscience

Huang J, Jung JY, Nam C. Estimating effective connectivity in Alzheimer’s disease progression: A dynamic causal modeling study.Frontiers in Human Neuroscience. 2022;16:no. 1060936

work page 2022
[77]

ParkHJ,PaeC,FristonK,JangC,RaziA,ZeidmanP,etal. Hierarchicaldynamiccausal modeling of resting-state fMRI reveals longitudinal changes in effective connectivity in the motor system after thalamotomy for essential tremor.Frontiers in Neurology. 2017;8:no. 346

work page 2017
[78]

Deconvolution of impulse response in event-related BOLD fMRI.NeuroImage

Glover G. Deconvolution of impulse response in event-related BOLD fMRI.NeuroImage. 1999;9:416-429

work page 1999
[79]

The effects of changes in PaCO2 on cerebral blood volume, blood flow, and vascular mean transit time.Stroke

Grubb R, Raichle M, Eichling J, Ter-Pogossian M. The effects of changes in PaCO2 on cerebral blood volume, blood flow, and vascular mean transit time.Stroke. 1974;5:630- 639

work page 1974
[80]

Oppenheim A, Willsky A, Nawab S.Signals and Systems. 2nd ed. Upper Saddle River, NJ: Prentice Hall; 1996

work page 1996

Showing first 80 references.