Enhanced inference of network structure from functional connectivity - - PowerPoint PPT Presentation

▶
enhanced inference of network structure from functional
SMART_READER_LITE
LIVE PREVIEW

Enhanced inference of network structure from functional connectivity - - PowerPoint PPT Presentation

Jun 08, 2023 236 likes •510 views

Enhanced inference of network structure from functional connectivity Eugene Duff Eugene Duff, Steve Smith, Tamar Makin, Mark Woolrich Brain connectivity in functional neuroimaging Cole, Nature Neuroscience, 2013 Brain connectivity in

slide-1
SLIDE 1

Enhanced inference of network structure from functional connectivity

Eugene Duff Eugene Duff, Steve Smith, Tamar Makin, Mark Woolrich

slide-2
SLIDE 2

Brain connectivity in functional neuroimaging

Cole, Nature Neuroscience, 2013

slide-3
SLIDE 3

Brain connectivity in functional neuroimaging

Correlation (network matrices) PPI Mutual Information Power/frequency analysis Coherence Dynamic Causal modelling Structural Equation modelling Granger Graph models Partial correlation

slide-4
SLIDE 4

Brain connectivity in functional neuroimaging

Correlation (network matrices) PPI Mutual Information Power/frequency analysis

Pros:

  • Can estimate more biologically

meaningful variables

  • Can model out uninteresting variables

Cons:

  • Often overparameterised, can only

compare limited set of model structures

  • Difficult to estimate for many nodes
  • Strong assumptions for ROIs, connectivity

patterns, signal dynamics, noise sources

  • All parameters/regions interdependent

Pros:

  • Simple pairwise measures permit,

comprehensive characterisation.

  • Computationally efficient.

Cons:

  • Can be sensitive to a wide range of effects:

ambiguous interpretation

  • Do not estimate neurophysiological

parameters

Dynamic Causal modelling Structural Equation modelling Granger Graph models Partial correlation Coherence How to use FC to inform EC model structure?

slide-5
SLIDE 5

.

What causes changes in correlations?

Y X

Healthy Patients

With no model of signal, analyses will be sensitive to variations in noise:

Y X

state 1 state 2

  • Increase in noise level

e.g. Differences in patient group, due to vascular tone.

  • Reduction in signal

e.g. Differences across activation states due to BOLD ceiling effect 𝜍x,y 𝜍x,y

Typical assumption is some sort of change in level of interaction between brain regions:

  • Strength of interaction
  • Extent of interaction
  • Activity level of network

𝜍x,y

slide-6
SLIDE 6

.

Aims

To identify a simple approach to characterising functional connectivity:

  • that is less ambiguous
  • remains a bivariate function for mapping and flexible implementation
  • provides insight into effective connectivity network modelling methods

(e.g. DCM, causal methods)

slide-7
SLIDE 7

Signal changes affecting correlations

𝜍x,y

Synchronisation of signals across regions

Y X

𝜍x,y

Increase in signal levels

𝜍x,y

Decrease in unrelated activity in one of two regions

𝜍x,y

Increase in common activity in one of two regions

𝜍x,y

Decrease in noise

𝜍x,y

Opposite changes in amplitude (combinations of above)

Y

X x y 𝜍x,y

Region 1 Region 2 BOLD signal

slide-8
SLIDE 8

Y X

𝜍x,y 𝜍x,y

Decrease in noise

𝜍x,y

Synchronisation of signals across regions

𝜍x,y

Increase in signal levels Decrease in unrelated activity in one of two regions

𝜍x,y

Increase in common activity in one of two regions

𝜍x,y

Opposite changes in amplitude (combinations of above)

Y

X x y 𝜍x,y

Region 1 Region 2 BOLD signal

σy

σx σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

Signal changes affecting correlations

slide-9
SLIDE 9

Potential dynamics affecting connectivity

By observing variances, we can eliminate the possibility of certain changes..

𝜍x,y

σy

σx

𝜍xb,b

σyb

σxb

Condition 1 Condition 2

Decrease in noise Synchronisation of signals across regions Increase in shared signal levels Decrease in unrelated activity in

  • ne of two regions

Increase in common activity in

  • ne of two regions

Opposite changes in amplitude

?

𝜍x,y σx,σxb σy, σyb

a < 𝜍xb,yb < b But the necessary extent of a change required to explain a specific change in correlation is unclear. Can we determine if certain variance changes explain a particular

  • bserved change?
slide-10
SLIDE 10

Shared/Unshared Signal Model

Proportion of shared signal in each region is bounded by correlation

Y

X

x y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2 ya w2 xa

𝜍xa,ya σya σxa

BOLD statistics (corr,var) proportion of shared signal y

Condition a

BOLD signal for condition a as a function of S and Ux Uy.

Model Formulation Pairwise model linking regions X and Y.

Y

X

x

y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2yb w2xb

𝜍xb,yb σyb σxb

BOLD statistics (corr,var) proportion of shared signal (y)

Condition b

slide-11
SLIDE 11

Shared/Unshared Signal Model

Y

X

x y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2 ya w2 xa

𝜍xb,yb σya σxa

BOLD statistics (corr,var) proportion of shared signal y

Condition a

Pairwise model linking regions X and Y.

Condition b produces some change in levels of shared and unshared signals - wxb = cx wx, wyb = cy wy. , matching the total change in variance: New observed variance can be expressed: And correlation:

Y

X

x

y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2yb w2xb

𝜍xb,yb σyb σxb

BOLD statistics (corr,var) proportion of shared signal (y)

Condition b

slide-12
SLIDE 12

Shared/Unshared Signal Model

Y

X

x y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2 ya w2 xa

𝜍xb,yb σya σxa

BOLD statistics (corr,var) proportion of shared signal y

Condition a

Pairwise model linking regions X and Y.

Y

X

x

y

Ux Uy S

BOLD signal Unshared signal (y) Unshared signal (x) Shared signal

w2yb w2xb

𝜍xb,yb σyb σxb

BOLD statistics (corr,var) proportion of shared signal (y)

Condition b

Given the limits on way, the effects of particular changes in signal and noise on 𝜍xb,yb can be determined based on variance changes. E.g. if there was no change in signal levels (variance was associated with unshared signal):

slide-13
SLIDE 13

𝜍x,y

Increase in signal levels

𝜍x,y

Increase in common activity in one of two regions

Identifying bounds on correlation

𝜍x,y 𝜍x,y

Decrease in noise

𝜍x,y

Synchronisation of signals across regions Decrease in unrelated activity in one of two regions

𝜍x,y

Opposite changes in amplitude (combinations of above)

σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

cx>1 cy>1 (ux=1) (uy=1) cx>1 cy=1 (ux=1) (uy=1) cy>1 cx>1 (uy<1) (ux<1) cx=1 cy=1 (ux<1) (uy<1) cx=1 cy=1 (ux=1) (uy<1) cy<1 cx>1 (uy=1) (ux=1)

slide-14
SLIDE 14

𝜍x,y

Increase in signal levels

𝜍x,y

Increase in common activity in one of two regions

Potential dynamics affecting connectivity

𝜍x,y 𝜍x,y

Decrease in noise

𝜍x,y

Synchronisation of signals across regions Decrease in unrelated activity in one of two regions

𝜍x,y

Opposite changes in amplitude (combinations of above)

σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

σx

σy

cx>1 cy>1 (ux=1) (uy=1) cx>1 cy=1 (ux=1) (uy=1) cx=1 cy=1 (ux>1) (uy>1) cx=1 cy=1 (ux>1) (uy=1) cy>1 cx>1 (uy<1) (ux<1) cy>1 cx<1 (uy=1) (ux=1)

slide-15
SLIDE 15

Active-state experiment

16 Subjects Multiband-6 TR: 1.3, 2x2x2mm voxels Five active-state conditions produce robust changes in ongoing dynamics providing a good testing set for connectivity measures.

slide-16
SLIDE 16

Active-state experiment

Are changes in connectivity across associated with variance changes? Do these changes correspond to particular types of changes in connectivity

  • are they predicted by model?
slide-17
SLIDE 17

Activation, variance, and correlation changes

Connectivity change from rest for visual condition (V1 seed) Connectivity change from rest for visual condition (V1 seed) Visual task activation Motor task activation Variance changes from rest for visual condition Variance changes from rest for motor condition

slide-18
SLIDE 18

Activation, variance, and correlation changes

Change in correlation (z-score)

Proportion Proportion of edges with at least

  • ne region changing in variance

Number of edges

Variance changes from rest for motor Variance changes from rest for visual

slide-19
SLIDE 19

Representing connectivity changes

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model

Cing Cuneus IntraCalc L a t O c c S u p D LingualD O c c P

  • l

e D PreCunD T e m p O c c CbL C b R PutL P u t R InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR L a t O c c S L LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL O c c P

  • l

e L OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus IntraCalc L a t O c c S u p D LingualD O c c P

  • l

e D PreCunD T e m p O c c CbL C b R PutL P u t R InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR L a t O c c S L LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL O c c P

  • l

e L OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus IntraCalc L a t O c c S u p D LingualD O c c P

  • l

e D PreCunD T e m p O c c CbL C b R PutL P u t R InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR L a t O c c S L LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL O c c P

  • l

e L OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus IntraCalc L a t O c c S u p D LingualD O c c P

  • l

e D PreCunD T e m p O c c CbL C b R PutL P u t R InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR L a t O c c S L LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL O c c P

  • l

e L OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model

Cing C u n e u s I n t r a C a l c LatOccSupD LingualD O c c P

  • l

e D PreCunD TempOcc CbL CbR PutL P u t R InfPreGL SMA I n f P r e G R P

  • s

t G L 2 PostGR2 P r e H L P r e H R LatOccL LatOccR L a t O c c S L LatOccSR LingActL LingActR OccFusAR OccFusAL O c c P

  • l

e L OccPoleR TempOccL T e m p O c c R L a t O c c I n f D

DMN Motor Visual

Change in shared signal

1

Cing C u n e u s I n t r a C a l c LatOccSupD LingualD O c c P

  • l

e D PreCunD TempOcc CbL CbR PutL P u t R InfPreGL SMA I n f P r e G R P

  • s

t G L 2 PostGR2 P r e H L P r e H R LatOccL LatOccR L a t O c c S L LatOccSR LingActL LingActR OccFusAR OccFusAL O c c P

  • l

e L OccPoleR TempOccL T e m p O c c R L a t O c c I n f D

DMN Motor Visual

Change in unshared signal

Cing C u n e u s I n t r a C a l c LatOccSupD LingualD O c c P

  • l

e D PreCunD TempOcc CbL CbR PutL P u t R InfPreGL SMA I n f P r e G R P

  • s

t G L 2 PostGR2 P r e H L P r e H R LatOccL LatOccR L a t O c c S L LatOccSR LingActL LingActR OccFusAR OccFusAL O c c P

  • l

e L OccPoleR TempOccL T e m p O c c R L a t O c c I n f D

DMN Motor Visual

Change in both shared and unshared signals

Cing C u n e u s I n t r a C a l c LatOccSupD LingualD O c c P

  • l

e D PreCunD TempOcc CbL CbR PutL P u t R InfPreGL SMA I n f P r e G R P

  • s

t G L 2 PostGR2 P r e H L P r e H R LatOccL LatOccR L a t O c c S L LatOccSR LingActL LingActR OccFusAR OccFusAL O c c P

  • l

e L OccPoleR TempOccL T e m p O c c R L a t O c c I n f D

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes Connections

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus IntraCalc LatOccSupD LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL SMA InfPreGR PostGL2 PostGR2 PreHL PreHR LatOccL LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR TempOccL TempOccR LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

slide-20
SLIDE 20

Changes between visual condition and rest

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus IntraCalc L a t O c c S u p D LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus IntraCalc L a t O c c S u p D LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus IntraCalc L a t O c c S u p D LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus IntraCalc L a t O c c S u p D LingualD OccPoleD PreCunD TempOcc CbL CbR PutL PutR InfPreGL S M A InfPreGR PostGL2 PostGR2 PreHL PreHR L a t O c c L LatOccR LatOccSL LatOccSR LingActL LingActR OccFusAR OccFusAL OccPoleL OccPoleR T e m p O c c L TempOccR LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

Visual stimulus related connectivity changes

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

Change in Shared Signal Change in Unshared Signal Change in Shared Signal Changes that can’t be explained by SNR changes Variance changes from rest for visual condition

slide-21
SLIDE 21

Changes between visual condition and rest

Visual task - Rest (V1 seed)

Changes indicative of decoupling

Significant reductions in in correlation

Correlation changes from rest for visual

Reduction in shared signal

Variance changes from rest for visual condition

slide-22
SLIDE 22

Changes between visual condition and rest

Visual task - Rest (V1 seed)

Changes indicative of decoupling

Significant reductions in in correlation

Correlation changes from rest for visual

Decoupling

Variance changes from rest for visual condition

slide-23
SLIDE 23

Changes between motor condition and rest

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

Change in Shared Signal Change in Unshared Signal Change in Shared Signal Changes that can’t be explained by SNR changes

Motor vs Rest conditions

Variance changes from rest for motor condition Motor stimulus related connectivity changes

Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes Correlation

  • B. Distribution of correlation changes
  • A. Correlation and variance changes between Rest and Visual Conditions categorised according to correlation-variance model
Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in shared signal

1

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in unshared signal

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in both shared and unshared signals

Cing Cuneus I n t r a C a l c L a t O c c S u p D LingualD OccPoleD PreCunD T e m p O c c CbL C b R PutL PutR InfPreGL S M A I n f P r e G R PostGL2 PostGR2 PreHL P r e H R L a t O c c L LatOccR LatOccSL LatOccSR L i n g A c t L LingActR O c c F u s A R OccFusAL OccPoleL OccPoleR T e m p O c c L T e m p O c c R LatOccInfD

DMN Motor Visual

Change in synchronisation Resting correlations Visual condition correlations Range of visual correlations that could be explained by variance changes

slide-24
SLIDE 24

Relation to network modelling (e.g. DCM)

Z Y X

+

Z X Y

+ +

Q Z X Y

+ +

slide-25
SLIDE 25

Relation to network modelling (e.g. DCM)

Z X Y

+ +

Decrease in noise Synchronisation of signals across regions Increase in shared signal levels Decrease in unrelated activity in

  • ne of two regions

Increase in common activity in

  • ne of two regions

Opposite changes in amplitude

Z X Y

+ +

Z X Y

+ +

slide-26
SLIDE 26

.

Summary

We have identified a simple model that links correlation and variance to provides insight into the types of dynamics underlying connectivity changes In a test dataset we could find almost every proposed feature of dynamics Most changes in correlation are accompanied by some change in variance DCM models typically predict variance changes, so are validated by these results Software is in development

Future directions

Application to population studies (HCP,DHCP) - cross subject variance Smooth integration with functional connectivity and DCM analyses More signal components: relationship to ICA?