Dynamic causal modeling: Difference between revisions

Revision as of 11:33, 14 August 2018

Latest revision as of 05:26, 15 August 2025

(43 intermediate revisions by 30 users not shown)

Line 1:

{{short description|Statistical modeling framework}}

{{AFC submission|d|cv|u=Peterzlondon|ns=118|decliner=Catrìona|declinets=20180813200742|ts=20180626194142}} <!-- Do not remove this line! -->

'''Dynamic causal modeling''' ('''DCM''') is a framework for specifying models, fitting them to data and comparing their evidence using [[Bayes factor|Bayesian model comparison]]. TheIt models, referred to as dynamic causal models (DCMs), areuses nonlinear [[State space|state-space]] models in continuous time, specified using [[Stochastic differential equation|stochastic]] or [[Ordinaryordinary differential equation|ordinary differential equations]]s. DCM was initially developed for identifyingtesting modelshypotheses ofabout [[Dynamical system|neural dynamics]].<ref name=":2Friston 2003">{{Cite journal|lastlast1=Friston|firstfirst1=K.J.|last2=Harrison|first2=L.|last3=Penny|first3=W.|date=August 2003-08|title=Dynamic causal modelling|url=https://doi.org/10.1016/S1053-8119(03)00202-7|journal=NeuroImage|volume=19|issue=4|pages=1273–1302|doi=10.1016/s1053-8119(03)00202-7|pmid=12948688|s2cid=2176588|issn=1053-8119}}</ref>. In this setting, differential equations describe the interaction of neural populations, which directly or indirectly give rise to functional neuroimaging data e.g., [[functional magnetic resonance imaging]] (fMRI), [[magnetoencephalography]] (MEG) or [[electroencephalography]] (EEG). Parameters in these DCMsmodels quantify the directed influences or effective connectivity among neuronal populations, which can beare estimated from the available data using [[Bayesian inference|Bayesian]] statistical methods.

<!-- EDIT BELOW THIS LINE -->

'''Dynamic causal modeling''' (DCM) is a framework for specifying models, fitting them to data and comparing their evidence using [[Bayes factor|Bayesian model comparison]]. The models, referred to as dynamic causal models (DCMs), are nonlinear [[State space|state-space]] models in continuous time, specified using [[Stochastic differential equation|stochastic]] or [[Ordinary differential equation|ordinary differential equations]]. DCM was initially developed for identifying models of [[Dynamical system|neural dynamics]]<ref name=":2">{{Cite journal|last=Friston|first=K.J.|last2=Harrison|first2=L.|last3=Penny|first3=W.|date=2003-08|title=Dynamic causal modelling|url=https://doi.org/10.1016/S1053-8119(03)00202-7|journal=NeuroImage|volume=19|issue=4|pages=1273–1302|doi=10.1016/s1053-8119(03)00202-7|issn=1053-8119}}</ref>. In this setting, differential equations describe the interaction of neural populations, which directly or indirectly give rise to functional neuroimaging data e.g., [[functional magnetic resonance imaging]] (fMRI), [[magnetoencephalography]] (MEG) or [[electroencephalography]] (EEG). Parameters in these DCMs quantify the directed influences or effective connectivity among neuronal populations, which can be estimated from the available data using [[Bayesian inference|Bayesian]] statistical methods.

== Procedure ==

DCM is usuallytypically used to estimate the coupling among brain regions and the changes in coupling due to experimental changes (e.g., time or context). ModelsA model of interacting brainneural regionspopulations areis specified, whichwith describea level of biological detail dependent on the interactionhypotheses ofand neuralavailable populationsdata. TheseThis areis supplementedcoupled with a forward model ofdescribing how the hidden states of each brain region (e.g., neuronalneural activity) givegives rise to measured responses. Having specified This enablesEstimating the bestgenerative model(s) andidentifies theirthe parameters (i.e.g. effectiveconnection connectivitystrengths) tofrom be identified fromthe observed data. [[Bayesian model comparison]] is used to compare models based on their evidence, which can then be characterised in terms of parameters (e.g. connection strengths).

DCM studies typically involve the following stages :<ref name=":0Stephan 2010">{{Cite journal|lastlast1=Stephan|firstfirst1=K.E.|last2=Penny|first2=W.D.|last3=Moran|first3=R.J.|author3-link=Rosalyn Moran|last4=den Ouden|first4=H.E.M.|last5=Daunizeau|first5=J.|last6=Friston|first6=K.J.|date=February 2010-02|title=Ten simple rules for dynamic causal modeling|url=https://dx.doi.org/10.1016/j.neuroimage.2009.11.015|journal=NeuroImage|volume=49|issue=4|pages=3099–3109|doi=10.1016/j.neuroimage.2009.11.015|pmid=19914382|pmc=2825373|issn=1053-8119}}</ref>:

# Experimental design. Specific hypotheses are formulated and an experiment is conducted.

Line 13 ⟶ 11:

# Model specification. One or more forward models (DCMs) are specified for each dataset.

#Model estimation. The model(s) are fitted to the data to determine their evidence and parameters.

# Model comparison. The evidence for each model is used for Bayesian Model Comparison (at the single-subject level or at the group level) orto select the best model(s). Bayesian model averaging (BMA) is used to providecompute a weighted average of parameter estimates, averaged over different models.

The key stepsstages are briefly reviewed below.

== Experimental design ==

Functional neuroimaging experiments are typically either task-based or examine brain activity at rest ([[Resting state fMRI|resting state]]). In task-based experiments, brain responses are evoked by known deterministic inputs (experimentally controlled stimuli) that embody designed changes in sensory stimulation or cognitive set. These experimental or exogenous variables can change neural activity in one of two ways. First, they can elicit responses through direct influences on specific brain regions. This would include, forsuch example,as [[Evokedevoked potential|sensory evoked responses]]s in the early visual cortex., Theor second class of inputs exerts their effects vicariously, throughvia a modulation of coupling among nodesneural populations; for example, the influence of attention on the processing of sensory information. These two types of input - driving and modulatory - are parameterized separately in DCM.<ref name=":2Friston 2003" />. To enable efficient estimation of driving and modulatory effects, a 2x2 [[Factorial experiment|factorial experimental design]] is often used - with one factor serving as the driving input and the other as the modulatory input .<ref name=":0Stephan 2010" />.

Resting state experiments have no experimental manipulations within the period of the neuroimaging recording. Instead, hypotheses are tested about the coupling of endogenous fluctuations in neuronal activity, or in the differences in connectivity between sessions or subjects. The DCM framework includes models and procedures for analysing resting state data, described belowin the next section.

== Model specification ==

All models in DCM have the following basic form:

Dynamic Causal Models (DCMs) are nonlinear state-space [[Dynamical system|dynamical with models]] in continuous time, parameterized in terms of directed effective connectivity between brain regions. Unlike [[Bayesian network|Bayesian Networks]], DCMs can be cyclic, and unlike [[Structural equation modeling|Structural Equation modelling]] and [[Granger causality]], DCM does not depend on the theory of Martingales; i.e., it does not assume that random fluctuations' are serially uncorrelated. All models in DCM have the following basic form:

<math>\begin{align}

Line 30 ⟶ 28:

\end{align}</math>

The first equality describes the change in neural activity <math>z</math> with respect to time (i.e. <math>\dot{z}</math>), which cannot be directly observed using non-invasive functional imaging modalities. The evolution of neural activity over time is controlled by a neural function <math>f</math> with parameters <math>\theta^{(n)}</math> and experimental inputs <math>u</math>. The neural activity in turn causes the timeseries <math>y</math>, (second equality)., The timesserieswhich are generated via aan observation function <math>g</math> with parameters <math>\theta^{(h)}</math>. Additive observation noise <math>\epsilon</math> completes the observation model. Usually, the key parameters of interest are the neural parameters <math>\theta^{(n)}</math> whichare of key interest, which for example, represent connection strengths that may change under different experimental conditions.

Specifying a DCM requires selecting a neural model <math>f</math> and observation model <math>g</math> and setting appropriate [[Prior probability|priors]] over the parameters; e.g. selecting which connections should be switched on or off.

==== Functional MRI ====

[[File:DCM for fMRI.svg|alt=DCM for fMRI neural circuit|thumb|The neural model in DCM for fMRI. z1 and z2 are the mean levels of activity in each region. Parameters A are the effective connectivity, B is the modulation of connectivity by a specific experimental condition and C is the driving input. ]]

The neural model in DCM for fMRI is a [[Taylor series|Taylor approximation]] that captures the gross causal influences between brain regions and their change due to experimental inputs (see picture). This is coupled with a detailed biophysical model of the generation of the blood oxygen level dependent (BOLD) response and the MRI signal,<ref name=":2">{{CiteFriston journal|last=Friston|first=K.J.|last2=Harrison|first2=L.|last3=Penny|first3=W.|date=2003-08|title=Dynamic causal modelling|url=https:"//doi.org/10.1016/S1053-8119(03)00202-7|journal=NeuroImage|volume=19|issue=4|pages=1273–1302|doi=10.1016/s1053-8119(03)00202-7|issn=1053-8119}}</ref>, based on the Balloon model of Buxton et al.,<ref>{{Cite journal|lastlast1=Buxton|firstfirst1=Richard B.|last2=Wong|first2=Eric C.|last3=Frank|first3=Lawrence R.|date=June 1998-06|title=Dynamics of blood flow and oxygenation changes during brain activation: The balloon model|url=https://dx.doi.org/10.1002/mrm.1910390602|journal=Magnetic Resonance in Medicine|volume=39|issue=6|pages=855–864|doi=10.1002/mrm.1910390602|issn=0740-3194|pmid=9621908|s2cid=2002497}}</ref>, which was supplemented with a model of neurovascular coupling .<ref>{{Cite journal|lastlast1=Friston|firstfirst1=K.J.|last2=Mechelli|first2=A.|last3=Turner|first3=R.|last4=Price|first4=C.J.|date=October 2000-10|title=Nonlinear Responses in fMRI: The Balloon Model, Volterra Kernels, and Other Hemodynamics|url=https://dx.doi.org/10.1006/nimg.2000.0630|journal=NeuroImage|volume=12|issue=4|pages=466–477|doi=10.1006/nimg.2000.0630|pmid=10988040|s2cid=961661|issn=1053-8119}}</ref><ref>{{Cite journal|lastlast1=Stephan|firstfirst1=Klaas Enno|last2=Weiskopf|first2=Nikolaus|last3=Drysdale|first3=Peter M.|last4=Robinson|first4=Peter A.|last5=Friston|first5=Karl J.|date=November 2007-11|title=Comparing hemodynamic models with DCM|url=https://dx.doi.org/10.1016/j.neuroimage.2007.07.040|journal=NeuroImage|volume=38|issue=3|pages=387–401|doi=10.1016/j.neuroimage.2007.07.040|pmid=17884583|pmc=2636182|issn=1053-8119}}</ref>. Additions to the neural model have included interactions between excitatory and inhibitory neural populations <ref>{{Cite journal|lastlast1=Marreiros|firstfirst1=A.C.|last2=Kiebel|first2=S.J.|last3=Friston|first3=K.J.|date=January 2008-01|title=Dynamic causal modelling for fMRI: A two-state model|url=https://doi.org/10.1016/j.neuroimage.2007.08.019|journal=NeuroImage|volume=39|issue=1|pages=269–278|doi=10.1016/j.neuroimage.2007.08.019|pmid=17936017|issn=1053-8119|citeseerx=10.1.1.160.1281|s2cid=9731930}}</ref> and non-linear influences of neural populations on the coupling between other populations.<ref name=":3Stephan 2008">{{Cite journal|lastlast1=Stephan|firstfirst1=Klaas Enno|last2=Kasper|first2=Lars|last3=Harrison|first3=Lee M.|last4=Daunizeau|first4=Jean|last5=den Ouden|first5=Hanneke E.M.|last6=Breakspear|first6=Michael|last7=Friston|first7=Karl J.|date=August 2008-08|title=Nonlinear dynamic causal models for fMRI|url=https://doi.org/10.1016/j.neuroimage.2008.04.262|journal=NeuroImage|volume=42|issue=2|pages=649–662|doi=10.1016/j.neuroimage.2008.04.262|issn=1053-8119|pmc=2636907|pmid=18565765}}</ref>.

DCM for resting state studies was first introduced in Stochastic DCM,<ref>{{Cite journal|date=2011-09-15|title=Generalised filtering and stochastic DCM for fMRI|url= https://www.sciencedirectzora.comuzh.ch/scienceid/articleeprint/pii49235/S10538119110014061/Li_Neuroimage_2011.pdf|journal=NeuroImage|language=en|volume=58|issue=2|pages=442–457|doi=10.1016/j.neuroimage.2011.01.085|pmid=21310247|issn=1053-8119|last1=Li|first1=Baojuan|last2=Daunizeau|first2=Jean|last3=Stephan|first3=Klaas E|last4=Penny|first4=Will|last5=Hu|first5=Dewen|last6=Friston|first6=Karl|s2cid=13956458}}</ref>, which estimates both neural fluctuations and connectivity parameters in the time ___domain, using [[Generalized filtering|Generalized Filtering]]. A more efficient scheme for resting state data was subsequently introduced which operates in the frequency ___domain, called DCM for Cross-Spectral Density (CSD) .<ref>{{Cite journal|lastlast1=Friston|firstfirst1=Karl J.|last2=Kahan|first2=Joshua|last3=Biswal|first3=Bharat|last4=Razi|first4=Adeel|date=July 2014-07|title=A DCM for resting state fMRI|url=https://dx.doi.org/10.1016/j.neuroimage.2013.12.009|journal=NeuroImage|volume=94|issue=100 |pages=396–407|doi=10.1016/j.neuroimage.2013.12.009|pmid=24345387|pmc=4073651|issn=1053-8119}}</ref><ref>{{Cite journal|lastlast1=Razi|firstfirst1=Adeel|last2=Kahan|first2=Joshua|last3=Rees|first3=Geraint|last4=Friston|first4=Karl J.|date=February 2015-02|title=Construct validation of a DCM for resting state fMRI|url=https://doi.org/10.1016/j.neuroimage.2014.11.027|journal=NeuroImage|volume=106|pages=1–14|doi=10.1016/j.neuroimage.2014.11.027|issn=1053-8119|pmc=4295921|pmid=25463471}}</ref>. Both of these can be applied to large-scale brain networks by constraining the connectivity parameters based on the functional connectivity.<ref>{{Cite journal|lastlast1=Seghier|firstfirst1=Mohamed L.|last2=Friston|first2=Karl J.|date=March 2013-03|title=Network discovery with large DCMs|url=https://doi.org/10.1016/j.neuroimage.2012.12.005|journal=NeuroImage|volume=68|pages=181–191|doi=10.1016/j.neuroimage.2012.12.005|issn=1053-8119|pmc=3566585|pmid=23246991}}</ref><ref name=":4Razi 2017">{{Cite journal|lastlast1=Razi|firstfirst1=Adeel|last2=Seghier|first2=Mohamed L.|last3=Zhou|first3=Yuan|last4=McColgan|first4=Peter|last5=Zeidman|first5=Peter|last6=Park|first6=Hae-Jeong|last7=Sporns|first7=Olaf|last8=Rees|first8=Geraint|last9=Friston|first9=Karl J.|date=October 2017-10|title=Large-scale DCMs for resting-state fMRI|url=https://doi.org/10.1162/NETN_a_00015|journal=Network Neuroscience|language=en|volume=1|issue=3|pages=222–241|doi=10.1162/netn_a_00015|issn=2472-1751|pmc=5796644|pmid=29400357}}</ref>. Another recent development for resting state analysis is Regression DCM<ref>{{Cite journal|lastlast1=Frässle|firstfirst1=Stefan|last2=Lomakina|first2=Ekaterina I.|last3=Razi|first3=Adeel|last4=Friston|first4=Karl J.|last5=Buhmann|first5=Joachim M.|last6=Stephan|first6=Klaas E.|date=July 2017-07|title=Regression DCM for fMRI|url=https://doi.org/10.1016/j.neuroimage.2017.02.090|journal=NeuroImage|volume=155|pages=406–421|doi=10.1016/j.neuroimage.2017.02.090|pmid=28259780|issn=1053-8119|doi-access=free|hdl=20.500.11850/182456|hdl-access=free}}</ref> implemented in the Tapas software collection (see [[#Software implementations|Software implementations]]). Regression DCM operates in the frequency ___domain, but linearizes the model under certain simplifications, such as having a fixed (canonical) haemodynamic response function. The enables rapid estimation of large-scale brain networks.

[[File:DCM for ERP and CMC.svg|thumb|Models of the cortical column used in EEG/MEG/LFP analysis. Self-connections on each population are present but not shown for clarity. Left: DCM for ERP. Right: Canonical Microcircuit (CMC). 1=spiny stellate cells (layer IV), 2=inhibitory interneurons, 3=(deep) pyramidal cells and 4=superficial pyramidal cells.]]

==== EEG / MEG ====

DCM for EEG and MEG data use more biologically detailed neural models than fMRI, asdue to the higher temporal resolution of these measurement techniques supports more refined models of neural dynamics. These can be classed into physiological models, which recapitulate neural circuitycircuitry, and phenomenological models, which focus on reproducing particular data features. The physiological models can be further subdivided into two classes. [http://www.scholarpedia.org/article/Conductance-based_models Conductance-based models] derive from the equivalent circuit representation of the cell membrane developed by Hodgkin and Huxley in the 1950s.<ref name=":5Hodgkin 1952">{{Cite journal|lastlast1=Hodgkin|firstfirst1=A. L.|last2=Huxley|first2=A. F.|date=1952-04-28|title=The components of membrane conductance in the giant axon ofLoligo|url=https://dx.doi.org/10.1113/jphysiol.1952.sp004718|journal=The Journal of Physiology|volume=116|issue=4|pages=473–496|doi=10.1113/jphysiol.1952.sp004718|pmid=14946714|issn=0022-3751|pmc=1392209}}</ref> . Convolution models were introduced by [[Wilson–Cowan model|Wilson & Cowan]]<ref>{{Cite journal|lastauthor2-link=Jack D. Cowan|last1=Wilson|firstfirst1=H. R.|last2=Cowan|first2=J. D.|date=September 1973-09|title=A mathematical theory of the functional dynamics of cortical and thalamic nervous tissue|url=https://dx.doi.org/10.1007/bf00288786|journal=Kybernetik|volume=13|issue=2|pages=55–80|doi=10.1007/bf00288786|pmid=4767470|s2cid=292546|issn=0340-1200}}</ref> and Freeman <ref>{{Cite journalbook|date=1975|title=Mass Action in the Nervous System|url=https://dx.doi.org/10.1016/c2009-0-03145-6|doi=10.1016/c2009-0-03145-6|isbn=9780122671500|last1=Freeman|first1=Walter J}}</ref> in the 1970s and involve a convolution of pre-synaptic input by a synaptic kernel function. TheSome of the specific models used in DCM are as follows:

* Physiological models:

** Convolution models:

*** DCM for evoked responses (DCM for ERP).<ref>{{Cite journal|lastlast1=David|firstfirst1=Olivier|last2=Friston|first2=Karl J.|date=November 2003-11|title=A neural mass model for MEG/EEG:|url=https://dx.doi.org/10.1016/j.neuroimage.2003.07.015|journal=NeuroImage|volume=20|issue=3|pages=1743–1755|doi=10.1016/j.neuroimage.2003.07.015|pmid=14642484|s2cid=1197179|issn=1053-8119}}</ref><ref>{{Citation|lastlast1=Kiebel|firstfirst1=Stefan J.|date=2009-07-31|url=https://dx.doi.org/10.7551/mitpress/9780262013086.003.0006|work=Brain Signal Analysis|pages=141–170|publisher=The MIT Press|isbn=9780262013086|last2=Garrido|first2=Marta I.|last3=Friston|first3=Karl J.|doi=10.7551/mitpress/9780262013086.003.0006|chapter=Dynamic Causal Modeling for Evoked Responses|title=Brain Signal Analysis}}</ref>. This is a biologically plausible neural mass model, extending earlier work by Jansen and Rit.<ref>{{Cite journal|lastlast1=Jansen|firstfirst1=Ben H.|last2=Rit|first2=Vincent G.|date=1995-09-01|title=Electroencephalogram and visual evoked potential generation in a mathematical model of coupled cortical columns|url=https://dx.doi.org/10.1007/s004220050191|journal=Biological Cybernetics|volume=73|issue=4|pages=357–366|doi=10.1007/s004220050191|pmid=7578475|issn=0340-1200}}</ref>. It emulates the activity of a cortical area using three neuronal sub-populations (see picture), each of which rests on two operators. The first operator transforms the pre-synaptic firing rate into a Post-Synaptic Potential (PSP), by [[Convolution|convolving]] pre-synaptic input with a synaptic response function (kernel). The second operator, a [[Sigmoid function|sigmoid]] function, transforms the membrane potential into a firing rate of action potentials.

*** DCM for LFP (Local Field Potentials).<ref>{{Cite journal|lastlast1=Moran|firstfirst1=R.J.|author3-link=Rosalyn Moran|last2=Kiebel|first2=S.J.|last3=Stephan|first3=K.E.|last4=Reilly|first4=R.B.|last5=Daunizeau|first5=J.|last6=Friston|first6=K.J.|date=September 2007-09|title=A neural mass model of spectral responses in electrophysiology|url=https://dx.doi.org/10.1016/j.neuroimage.2007.05.032|journal=NeuroImage|volume=37|issue=3|pages=706–720|doi=10.1016/j.neuroimage.2007.05.032|pmid=17632015|pmc=2644418|issn=1053-8119}}</ref>. Extends DCM for ERP by adding the effects of specific ion channels on spike generation.

*** Canonical Microcircuit (CMC).<ref>{{Cite journal|lastlast1=Bastos|firstfirst1=Andre  M.|last2=Usrey|first2=W.  Martin|last3=Adams|first3=Rick  A.|last4=Mangun|first4=George  R.|last5=Fries|first5=Pascal|last6=Friston|first6=Karl  J.|date=November 2012-11|title=Canonical Microcircuits for Predictive Coding|url=https://dx.doi.org/10.1016/j.neuron.2012.10.038|journal=Neuron|volume=76|issue=4|pages=695–711|doi=10.1016/j.neuron.2012.10.038|pmid=23177956|pmc=3777738|issn=0896-6273}}</ref>. Used to address hypotheses about laminar-specific ascending and descending connections in the brain, which underpin the [[predictive coding]] account of functional brain architectures. The single pyramidal cell population from DCM for ERP is split into deep and superficial populations (see picture). A version of the CMC has been applied to model multi-modal MEG and fMRI data.<ref>{{Cite journal|lastlast1=Friston|firstfirst1=K.J.|last2=Preller|first2=Katrin H.|last3=Mathys|first3=Chris|last4=Cagnan|first4=Hayriye|last5=Heinzle|first5=Jakob|last6=Razi|first6=Adeel|last7=Zeidman|first7=Peter|date=February 2017-02|title=Dynamic causal modelling revisited|url=https://doi.org/10.1016/j.neuroimage.2017.02.045|journal=NeuroImage|volume=199|pages=730–744|doi=10.1016/j.neuroimage.2017.02.045|pmid=28219774|pmc=6693530|issn=1053-8119}}</ref>.

***Neural Field Model (NFM).<ref>{{Cite journal|lastlast1=Pinotsis|firstfirst1=D.A.|last2=Friston|first2=K.J.|date=March 2011-03|title=Neural fields, spectral responses and lateral connections|url=https://dx.doi.org/10.1016/j.neuroimage.2010.11.081|journal=NeuroImage|volume=55|issue=1|pages=39–48|doi=10.1016/j.neuroimage.2010.11.081|pmid=21138771|pmc=3049874|issn=1053-8119}}</ref>. Extends the models above into the spatial ___domain, modelling continuous changes in current across the cortical sheet.

** Conductance models:

***Neural Mass Model (NMM) and Mean-field model (MFM).<ref>{{Cite journal|lastlast1=Marreiros|firstfirst1=André C.|last2=Daunizeau|first2=Jean|last3=Kiebel|first3=Stefan J.|last4=Friston|first4=Karl J.|date=August 2008-08|title=Population dynamics: Variance and the sigmoid activation function|url=https://dx.doi.org/10.1016/j.neuroimage.2008.04.239|journal=NeuroImage|volume=42|issue=1|pages=147–157|doi=10.1016/j.neuroimage.2008.04.239|pmid=18547818|s2cid=13932515|issn=1053-8119}}</ref><ref>{{Cite journal|lastlast1=Marreiros|firstfirst1=André C.|last2=Kiebel|first2=Stefan J.|last3=Daunizeau|first3=Jean|last4=Harrison|first4=Lee M.|last5=Friston|first5=Karl J.|date=February 2009-02|title=Population dynamics under the Laplace assumption|url=https://dx.doi.org/10.1016/j.neuroimage.2008.10.008|journal=NeuroImage|volume=44|issue=3|pages=701–714|doi=10.1016/j.neuroimage.2008.10.008|pmid=19013532|s2cid=12369912|issn=1053-8119}}</ref>. These have the same arrangement of neural populations as DCM for ERP, above, but are based on the [[Morris–Lecar model|Morris-Lecar model]] of the barnacle muscle fibre ,<ref>{{Cite journal|lastlast1=Morris|firstfirst1=C.|last2=Lecar|first2=H.|date=July 1981-07|title=Voltage oscillations in the barnacle giant muscle fiber|url=https://dx.doi.org/10.1016/s0006-3495(81)84782-0|journal=Biophysical Journal|volume=35|issue=1|pages=193–213|doi=10.1016/s0006-3495(81)84782-0|pmid=7260316|pmc=1327511|bibcode=1981BpJ....35..193M|issn=0006-3495}}</ref>, which in turn derives from the [[Hodgkin–Huxley model|Hodgin and Huxley]] model of the giant squid axon.<ref name=":5Hodgkin 1952" />. They enable inference about ligand-gated excitatory (Na+) and inhibitory (Cl-) ion flow, mediated through fast glutamatergic and GABAergic receptors. Whereas DCM for fMRI and the convolution models represent the activity of each neural population by a single number - its mean activity - the conductance models include the full density (probability distribution) of activity within the population. The 'mean-field assumption' used in the MFM version of the model assumes the density of one population's activity depends only on the mean of another. A subsequent extension added voltage-gated NMDA ion channels.<ref>{{Cite journal|lastlast1=Moran|firstfirst1=Rosalyn J.|author1-link=Rosalyn Moran|last2=Stephan|first2=Klaas E.|last3=Dolan|first3=Raymond J.|last4=Friston|first4=Karl J.|date=April 2011-04|title=Consistent spectral predictors for dynamic causal models of steady-state responses|url=https://doi.org/10.1016/j.neuroimage.2011.01.012|journal=NeuroImage|volume=55|issue=4|pages=1694–1708|doi=10.1016/j.neuroimage.2011.01.012|issn=1053-8119|pmc=3093618|pmid=21238593}}</ref>.

****

* Phenomenological models:

**DCM for phase coupling.<ref>{{Cite journal|lastlast1=Penny|firstfirst1=W.D.|last2=Litvak|first2=V.|last3=Fuentemilla|first3=L.|last4=Duzel|first4=E.|last5=Friston|first5=K.|date=September 2009-09|title=Dynamic Causal Models for phase coupling|url=https://dx.doi.org/10.1016/j.jneumeth.2009.06.029|journal=Journal of Neuroscience Methods|volume=183|issue=1|pages=19–30|doi=10.1016/j.jneumeth.2009.06.029|pmid=19576931|pmc=2751835|issn=0165-0270}}</ref>. Models the interaction of brain regions as Weakly Coupled Oscillators (WCOs), in which the rate of change of phase of one oscillator is related to the phase differences between itself and other oscillators.

== Model estimation ==

Model inversion or estimation is implemented in DCM using [[Variational Bayesian methods|variational Bayes]] under the [[Laplace's method|Laplace assumption]].<ref>{{Citation|lastlast1=Friston|firstfirst1=K.|date=2007|url=https://dx.doi.org/10.1016/b978-012372560-8/50047-4|work=Statistical Parametric Mapping|pages=606–618|publisher=Elsevier|isbn=9780123725608|last2=Mattout|first2=J.|last3=Trujillo-Barreto|first3=N.|last4=Ashburner|first4=J.|last5=Penny|first5=W.|doi=10.1016/b978-012372560-8/50047-4|chapter=Variational Bayes under the Laplace approximation|title=Statistical Parametric Mapping}}</ref>. This provides two useful quantities: the log marginal likelihood or model evidence <math>\ln{p(y|m)}</math> is the probability of observing of the data under a given model. Generally, this cannot be calculated explicitly and is approximated by a quantity called the negative variational free energy <math>F</math> , referred to in machine learning as the Evidence Lower Bound (ELBO). Hypotheses are tested by comparing the evidence for different models based on their free energy, a procedure called Bayesian model comparison.

Model estimation also provides estimates of the parameters <math>p(\theta|y)</math>;, for example, connection strengths, which maximise variationalthe free energy. Where models differ only in their priors, [[Bayesian model reduction|Bayesian Model Reduction]] can be used to derive the evidence and parameters of nested or reduced models analytically and efficiently.

== Model comparison ==

Neuroimaging studies typically investigate effects that are conserved at the group level, or which differ between subjects. There are two predominant approaches for group-level analysis: random effects Bayesian Model Selection (BMS)<ref>{{Cite journal|lastlast1=Rigoux|firstfirst1=L.|last2=Stephan|first2=K.E.|last3=Friston|first3=K.J.|last4=Daunizeau|first4=J.|date=January 2014-01|title=Bayesian model selection for group studies — Revisited|url=https://dx.doi.org/10.1016/j.neuroimage.2013.08.065|journal=NeuroImage|volume=84|pages=971–985|doi=10.1016/j.neuroimage.2013.08.065|pmid=24018303|s2cid=1908433|issn=1053-8119}}</ref> and Parametric Empirical Bayes (PEB).<ref name=":1Friston 2016">{{Cite journal|lastlast1=Friston|firstfirst1=Karl J.|last2=Litvak|first2=Vladimir|last3=Oswal|first3=Ashwini|last4=Razi|first4=Adeel|last5=Stephan|first5=Klaas E.|last6=van Wijk|first6=Bernadette C.M.|last7=Ziegler|first7=Gabriel|last8=Zeidman|first8=Peter|date=March 2016-03|title=Bayesian model reduction and empirical Bayes for group (DCM) studies|url=https://doi.org/10.1016/j.neuroimage.2015.11.015|journal=NeuroImage|volume=128|pages=413–431|doi=10.1016/j.neuroimage.2015.11.015|issn=1053-8119|pmc=4767224|pmid=26569570}}</ref>. Random effectsEffects BMS posits that subjects differ in terms of which model generated their data - e.g. drawing a random subject from the population, there might be a 25% chance that their brain is structured like model 1 and a 75% chance that it is structured like model 2. The analysis pipeline for the BMS approach procedure follows a series of steps:

# Specify and estimate multiple DCMs per subject, where each DCM (or set of DCMs) embodies a hypothesis.

# Perform randomRandom effectsEffects BMS to estimate the proportion of subjects whose data were generated by each model

# Calculate the average connectivity parameters across models using Bayesian Model Averaging. This average is weighted by the posterior probability for each model., This meansmeaning that models with greater probability contribute more to the average than models with lower probability.

Alternatively, oneParametric can use parametric empiricalEmpirical Bayes (PEB) <ref name=":1Friston 2016" /> to considercan be used, which specifies a hierarchical model over parameters (e.g., connection strengths). It eschews the notion of different models at the level of individual subjects, and assumes that people differ in the (parametric) strength of connections. The PEB approach models distinct sources of variability in connection strengths across subjects using fixed effects and between-subject variability (random effects). The PEB procedure is as follows:

# Specify a single 'full' DCM per subject, which contains all the parameters of interest.

# Specify a Bayesian [[General linear model|General Linear Model (GLM)]] to model the parameters (the full posterior density) from all subjects at the group level.

# Test hypotheses by comparing the full group-level model to reduced group-level models where certain combinations of connections have been switched off.

== Validation ==

Developments in DCM have been validated using different approaches:

* Face validity establishes whether the parameters of a model can be recovered from simulated data. This is usually performed alongside the development of each new model (E.g. <ref name=":2Friston 2003" /><ref name=":3Stephan 2008" />).

* Construct validity assesses consistency with other analytical methods. For example, DCM has been compared with Structural Equation Modelling <ref>{{Cite journal|lastlast1=Penny|firstfirst1=W.D.|last2=Stephan|first2=K.E.|last3=Mechelli|first3=A.|last4=Friston|first4=K.J.|date=January 2004-01|title=Modelling functional integration: a comparison of structural equation and dynamic causal models|url=https://dx.doi.org/10.1016/j.neuroimage.2004.07.041|journal=NeuroImage|volume=23|pages=S264–S274|doi=10.1016/j.neuroimage.2004.07.041|pmid=15501096|issn=1053-8119|citeseerx=10.1.1.160.3141|s2cid=8993497}}</ref> and other neurobiological computational models .<ref>{{Cite journal|lastlast1=Lee|firstfirst1=Lucy|last2=Friston|first2=Karl|last3=Horwitz|first3=Barry|date=May 2006-05|title=Large-scale neural models and dynamic causal modelling|url=https://dx.doi.org/10.1016/j.neuroimage.2005.11.007|journal=NeuroImage|volume=30|issue=4|pages=1243–1254|doi=10.1016/j.neuroimage.2005.11.007|pmid=16387513|s2cid=19003382|issn=1053-8119}}</ref>.

* Predictive validity assesses the ability to predict known or expected effects. This has included testing against iEEG / EEG / stimulation <ref>{{Cite journal|lastlast1=David|firstfirst1=Olivier|last2=Guillemain|first2=Isabelle|last3=Saillet|first3=Sandrine|last4=Reyt|first4=Sebastien|last5=Deransart|first5=Colin|last6=Segebarth|first6=Christoph|last7=Depaulis|first7=Antoine|date=2008-12-23|title=Identifying Neural Drivers with Functional MRI: An Electrophysiological Validation|url=http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.0060315|journal=PLOS Biology|language=en|volume=6|issue=12|pages=e3152683–97|doi=10.1371/journal.pbio.0060315|issn=1545-7885|pmc=2605917|pmid=19108604 |doi-access=free }}</ref><ref>{{Cite journal|lastlast1=David|firstfirst1=Olivier|last2=Woźniak|first2=Agata|last3=Minotti|first3=Lorella|last4=Kahane|first4=Philippe|date=February 2008-02|title=Preictal short-term plasticity induced by intracerebral 1  Hz stimulation|url=https://doi.org/10.1016/j.neuroimage.2007.11.005|journal=NeuroImage|volume=39|issue=4|pages=1633–1646|doi=10.1016/j.neuroimage.2007.11.005|pmid=18155929|s2cid=3415312|issn=1053-8119|url=https://www.hal.inserm.fr/inserm-00381199/file/David_Manuscript.pdf}}</ref><ref>{{Cite journal|lastlast1=Reyt|firstfirst1=Sébastien|last2=Picq|first2=Chloé|last3=Sinniger|first3=Valérie|last4=Clarençon|first4=Didier|last5=Bonaz|first5=Bruno|last6=David|first6=Olivier|date=October 2010-10|title=Dynamic Causal Modelling and physiological confounds: A functional MRI study of vagus nerve stimulation|url=https://dx.doi.org/10.1016/j.neuroimage.2010.05.021|journal=NeuroImage|volume=52|issue=4|pages=1456–1464|doi=10.1016/j.neuroimage.2010.05.021|pmid=20472074|s2cid=1668349|issn=1053-8119|url=https://www.hal.inserm.fr/inserm-00498678/file/Manuscript_Author.pdf}}</ref><ref>{{Cite journal|lastlast1=Daunizeau|firstfirst1=J.|last2=Lemieux|first2=L.|last3=Vaudano|first3=A. E.|last4=Friston|first4=K. J.|last5=Stephan|first5=K. E.|date=2013|title=An electrophysiological validation of stochastic DCM for fMRI|url=https://dx.doi.org/10.3389/fncom.2012.00103|journal=Frontiers in Computational Neuroscience|volume=6|pages=103|doi=10.3389/fncom.2012.00103|pmid=23346055|pmc=3548242|issn=1662-5188|doi-access=free}}</ref> and against known pharmacological treatments .<ref>{{Cite journal|lastlast1=Moran|firstfirst1=Rosalyn  J.|author1-link=Rosalyn Moran|last2=Symmonds|first2=Mkael|last3=Stephan|first3=Klaas  E.|last4=Friston|first4=Karl  J.|last5=Dolan|first5=Raymond  J.|date=August 2011-08|title=An In  Vivo Assay of Synaptic Function Mediating Human Cognition|url=https://dx.doi.org/10.1016/j.cub.2011.06.053|journal=Current Biology|volume=21|issue=15|pages=1320–1325|doi=10.1016/j.cub.2011.06.053|pmid=21802302|pmc=3153654|issn=0960-9822}}</ref><ref>{{Cite journal|lastlast1=Moran|firstfirst1=Rosalyn J.|author1-link=Rosalyn Moran|last2=Jung|first2=Fabienne|last3=Kumagai|first3=Tetsuya|last4=Endepols|first4=Heike|last5=Graf|first5=Rudolf|last6=Dolan|first6=Raymond J.|last7=Friston|first7=Karl J.|last8=Stephan|first8=Klaas E.|last9=Tittgemeyer|first9=Marc|date=2011-08-02|title=Dynamic Causal Models and Physiological Inference: A Validation Study Using Isoflurane Anaesthesia in Rodents|url=https://dx.doi.org/10.1371/journal.pone.0022790|journal=PLoSPLOS ONE|volume=6|issue=8|pages=e22790|doi=10.1371/journal.pone.0022790|pmid=21829652|pmc=3149050|bibcode=2011PLoSO...622790M|issn=1932-6203|doi-access=free}}</ref>

== Limitations / drawbacks ==

DCM is a hypothesis-driven approach for investigating the interactions among pre-defined regions of interest. It is not ideally suited for exploratory analyses.<ref name=":0Stephan 2010" />. Although methods have been implemented for automatically searching over reduced models ([[Bayesian model reduction|Bayesian Model Reduction]]) and for modelling large-scale brain networks,<ref name=":4Razi 2017" />, these methods require an explicit specification of model space. inIn neuroimaging, other approaches such as [[Psychophysiological Interaction| psychophysiological interaction (PPI)]] analysis may be more appropriate for exploratory use; especially for discovering key nodes for subsequent DCM analysis.

The variational Bayesian methods used for model estimation in DCM are based on the on the Laplace assumption, thatwhich treats the posterior over parameters isas Gaussian. This approximation can fail in the context of highly non-linear models, where local minima canmay preclude the free energy from serving as a tight bound on log model evidence. Sampling approaches provide the gold standard; however, they are time -consuming to run and have usuallytypically been used to validate the variational approximations in DCM.<ref>{{Cite journal|lastlast1=Chumbley|firstfirst1=Justin R.|last2=Friston|first2=Karl J.|last3=Fearn|first3=Tom|last4=Kiebel|first4=Stefan J.|date=November 2007-11|title=A Metropolis–Hastings algorithm for dynamic causal models|url=https://dx.doi.org/10.1016/j.neuroimage.2007.07.028|journal=NeuroImage|volume=38|issue=3|pages=478–487|doi=10.1016/j.neuroimage.2007.07.028|pmid=17884582|s2cid=3347682|issn=1053-8119}}</ref>

== Software implementations ==

DCM is implemented in the [[Statistical parametric mapping|Statistical Parametric Mapping]] software package, which serves as the canonical or reference implementation (http://www.fil.ion.ucl.ac.uk/spm/software/spm12/). It has been re-implemented and developed in the Tapas software collection (https://www.tnu.ethz.ch/en/software/tapas.html {{Webarchive|url=https://web.archive.org/web/20190203134649/https://www.tnu.ethz.ch/en/software/tapas.html |date=2019-02-03 }}) and the VBA toolbox (https://mbb-team.github.io/VBA-toolbox/).

== References ==

{{Reflist}}

== Further reading ==

{{Scholia}}

* [http://www.scholarpedia.org/article/Dynamic_causal_modeling Dynamic Causal Modelling on Scholarpedia]

* Understanding DCM: ten simple rules for the clinician<ref>{{Cite journal|last1=Kahan|first1=Joshua|last2=Foltynie|first2=Tom|date=December 2013|title=Understanding DCM: Ten simple rules for the clinician|journal=NeuroImage|volume=83|pages=542–549|doi=10.1016/j.neuroimage.2013.07.008|pmid=23850463|issn=1053-8119|doi-access=free}}</ref>

* Ten simple rules for dynamic causal modeling<ref name=":0" />

* UnderstandingNeural DCM:masses tenand simplefields rulesin fordynamic thecausal clinicianmodeling<ref>{{Cite journal|lastlast1=KahanMoran|firstfirst1=JoshuaRosalyn|author1-link=Rosalyn Moran|last2=FoltyniePinotsis|first2=TomDimitris A.|last3=Friston|first3=Karl|date=2013-12|title=UnderstandingNeural DCM:masses Tenand simplefields rulesin fordynamic thecausal clinician|url=https://dx.doi.org/10.1016/j.neuroimage.2013.07.008modeling|journal=NeuroImageFrontiers in Computational Neuroscience|volume=837|pages=542–54957|doi=10.10163389/j.neuroimagefncom.2013.07.00800057|pmid=23755005|pmc=3664834|issn=10531662-81195188|doi-access=free}}</ref>

* Neural masses and fields in dynamic causal modeling<ref>{{Cite journal|last=Moran|first=Rosalyn|last2=Pinotsis|first2=Dimitris A.|last3=Friston|first3=Karl|date=2013|title=Neural masses and fields in dynamic causal modeling|url=https://dx.doi.org/10.3389/fncom.2013.00057|journal=Frontiers in Computational Neuroscience|volume=7|doi=10.3389/fncom.2013.00057|issn=1662-5188}}</ref>

[[:Category:Neuroimaging]]

== References ==