Data and dynamic causal modelling Karl Friston Giuseppe Arcimboldo , - - PowerPoint PPT Presentation

Data and dynamic causal modelling

Karl Friston

Giuseppe Arcimboldo, The Vegetable Gardener (c.1590). Oil on panel. Our percepts are constrained by what we expect to see and the hypotheses that can be called upon to explain our sensory input. Arcimboldo, "a 16th century Milanese artist who was a favorite of the Viennese, illustrates this dramatically by using fruits and vegetables to create faces in his paintings.

Making sense of data ( ) ( , ) Q F s D   = −   = −    μ μ μ

prediction update

( ) s g  = − μ

prediction error

μ

Acquire MEG data and fMRI data Statistical parametric mapping DCM for electromagnetic responses Lamina specific (canonical microcircuit) DCM for fMRI fMRI data MEG data ECD priors

Multimodal DCM (Bayesian fusion)

Watching the brain make sense of data

Neuronal priors Bayesian belief updating

• r evidence accumulation

The forward (dynamic causal) model causes consequences

The forward (dynamic causal) model causes consequences Bayesian model inversion ( ) ( )

| , ( | ) ln | ( , ) p m q p m F         

Posterior density Log model evidence (Free energy) Richard Feynman

Dynamic causal modelling of synaptopathy

Karl Friston

The forward (dynamic causal) model

Positive

• Test for coronavirus

Death

• Recorded death

Susceptible Infected Infectious Immune

(2) infection

Untested

Waiting Positive Negative

(4) testing

Home Work CCU Morgue

(1) location (3) clinical

asymptomatic Symptom s

ARDS

Deceased

causes consequences