Linking brain systems and mental functions requires accurate descriptions of behavioral tasks and fine demarcations of brain regions. Functional Magnetic Resonance Imaging (fMRI) has contributed to the investigation of brain regions involved in a variety of cognitive processes. However, to date, no data collection has systematically addressed the functional mapping of cognitive mechanisms at a fine spatial scale. The Individual Brain Charting (IBC) project stands for a high-resolution multi-task fMRI dataset that intends to provide the objective basis toward a comprehensive functional atlas of the human brain. The data refer to a permanent cohort performing many different tasks. The large amount of task-fMRI data on the same subjects yields a precise mapping of the underlying functions, free from both inter-subject and inter-site variability. The first release of the IBC dataset consists of data acquired from thirteen participants during performance of a dozen of tasks. Raw data from this release are publicly available in the OpenNeuro repository and derived statistical maps can be found in NeuroVault [1]. These maps reveal a successful cognitive encoding of many psychological domains in large areas of the human brain. Indeed, main findings of the original studies were replicated at higher resolution. Our results thus provide a comprehensive revision of the neural correlates underlying behavior, highlighting nonetheless the spatial variability of functional signatures between participants. In addition, this dataset supports investigations using alternative approaches to group-level analysis of task-specific studies. For instance, such rich task-wise dataset can be applied to mega-analytic encoding models towards the development of a brain-atlasing framework, by systematically mapping functional signatures associated with the cognitive components of the tasks.

1. Functional specialization
in human cognition:
a large-scale neuroimaging initiative
Ana Lu´ısa Pinho
@ALuisaPinho
Parietal Team meeting
19th
of March 2019

2. Background and motivations
In cognitive neuroscience:
Brain systems
⇐⇒
Mental functions

3. Background and motivations
In cognitive neuroscience:
Brain systems
⇐⇒
Mental functions
tackle one psychological domain

4. Background and motivations
In cognitive neuroscience:
Brain systems
⇐⇒
Mental functions
tackle one psychological domain
be specific enough to accurately isolate brain
processes

5. Background and motivations
In cognitive neuroscience:
Brain systems
⇐⇒
Mental functions
tackle one psychological domain
be specific enough to accurately isolate brain
processes
⇓
Very hard to achieve!
Lack of generality.

6. Background and motivations
In cognitive neuroscience:
Brain systems
⇐⇒
Mental functions
Task-fMRI experiments allow to:
link brain systems to behavior
map neural activity at mm-scale

7. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)

8. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s

9. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s
Task-wise dataset:
Many tasks

10. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s
Task-wise dataset:
Many tasks
Fixed cohort - 13 healthy adults

11. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s
Task-wise dataset:
Many tasks
Fixed cohort - 13 healthy adults
Fixed environment
NeuroSpin platform, CEA-Saclay, France
Siemens 3T Magnetom Prismafit
64-channel coil

12. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s
Task-wise dataset:
Many tasks
Fixed cohort - 13 healthy adults
Fixed environment
Inclusion of other MRI modalities NeuroSpin platform, CEA-Saclay, France
Siemens 3T Magnetom Prismafit
64-channel coil

13. Features of the IBC dataset
High spatial-resolution fMRI data
(1.5mm)
TR = 2s
Task-wise dataset:
Many tasks
Fixed cohort - 13 healthy adults
Fixed environment
Inclusion of other MRI modalities
Not a longitudinal study!
NeuroSpin platform, CEA-Saclay, France
Siemens 3T Magnetom Prismafit
64-channel coil

14. Tasks
A task is:
a well-controlled sequence of behavioral operations

15. Tasks
A task is:
a well-controlled sequence of behavioral operations
Paradigms:
categorical designs ⇒ subtraction

16. Tasks
A task is:
a well-controlled sequence of behavioral operations
Paradigms:
categorical designs ⇒ subtraction
naturalistic stimuli

17. Tasks
A task is:
a well-controlled sequence of behavioral operations
Paradigms:
categorical designs ⇒ subtraction
naturalistic stimuli
Long-range cognitive order:
Perception ⇒ High-cognition

18. First release
Tasks
ARCHI tasks
Standard
Spatial
Social
Emotional
HCP tasks
Emotion
Gambling
Motor
Language
Relational
Social
Working Memory
RSVP Language
Sensory processing:
Retinotopy
Tonotopy
Somatotopy
High-cognitive order:
Calculation
Language
Social cognition
Theory-of-mind

19. First release
Tasks
ARCHI tasks
Standard
Spatial
Social
Emotional
HCP tasks
Emotion
Gambling
Motor
Language
Relational
Social
Working Memory
RSVP Language
Sensory processing:
Retinotopy
Tonotopy
Somatotopy
High-cognitive order:
Calculation
Language
Social cognition
Theory-of-mind
59 independent conditions

20. Behavioral protocols
Software Tools:
Presentation

21. Behavioral protocols
Software Tools:
Presentation
Protocols available on:
hbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocols

22. Behavioral protocols
Software Tools:
Presentation
Protocols available on:
hbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocolshbp-brain-charting/public protocols

23. Storage of raw MRI data
Link

24. Storage of raw MRI data
Link
ds000244
Link

25. Storage of raw MRI data
Link
ds000244
Link
Data organization: BIDS Specification

26. Analysis pipeline
Python package of the IBC analysis pipeline
available here:
hbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis codehbp-brain-charting/public analysis code

27. GLM univariate analysis

28. GLM univariate analysis
Y = β1X1+β2X2+

29. GLM univariate analysis
Finding ˆβ’s, so that
ˆY is as close as possible to Y .
ˆY = ˆβ1X1 + ˆβ2X2 Y = β1X1+β2X2+

30. GLM univariate analysis
Finding ˆβ’s, so that
ˆY is as close as possible to Y .
ˆY = ˆβ1X1 + ˆβ2X2 Y = β1X1+β2X2+
Univariate: fit the model to the time course of each voxel

31. GLM univariate analysis
Finding ˆβ’s, so that
ˆY is as close as possible to Y .
ˆY = ˆβ1X1 + ˆβ2X2 Y = β1X1+β2X2+
Univariate: fit the model to the time course of each voxel
Convolution with HRF: Xi = si (t) h(τ)

32. Contrasts and t-test
Contrasts: linear combinations of the β’s
Example:
cT
β = 1.β1 + 0.β2

33. Contrasts and t-test
Contrasts: linear combinations of the β’s
Example:
cT
β = 1.β1 + 0.β2
Hypothesis Testing ⇒ H0 : cT β = 0 Ha : cT β = 0

34. Contrasts and t-test
Contrasts: linear combinations of the β’s
Example:
cT
β = 1.β1 + 0.β2
Hypothesis Testing ⇒ H0 : cT β = 0 Ha : cT β = 0
Test statistic: summarizes evidence against the null hypothesis
Example:
t =
cT ˆβ
cT β
∼ tN−p

35. Contrasts and t-test
Contrasts: linear combinations of the β’s
Example:
cT
β = 1.β1 + 0.β2
Hypothesis Testing ⇒ H0 : cT β = 0 Ha : cT β = 0
Test statistic: summarizes evidence against the null hypothesis
Example:
t =
cT ˆβ
cT β
∼ tN−p
Access statistical significance corrected for multiple comparisons

36. Contrasts and t-test
Contrasts: linear combinations of the β’s
Example:
cT
β = 1.β1 + 0.β2
Hypothesis Testing ⇒ H0 : cT β = 0 Ha : cT β = 0
Test statistic: summarizes evidence against the null hypothesis
Example:
t =
cT ˆβ
cT β
∼ tN−p
Access statistical significance corrected for multiple comparisons
Hierarchical model: within- and between-subject analysis

37. Results

38. Post-processed data
Collection id=4438
Link
Group-level z-maps
qFDR < 0.05

39. Post-processed data
Collection id=4438
Link
Group-level z-maps
qFDR < 0.05

40. Conjunction provides better group-map consistency
RFX
smoothed
RFX
unsmoothed
Conjunction 25%
smoothed
Conjunction 25%
unsmoothed
0.5
0.6
0.7
0.8
0.9
Jaccardindex
Distribution of map consistency for all contrasts
between original set and bootstrap resampling

41. Brain coverage
Group-level F-map pFWE < 0.05
x=27x=12x=-13x=-28x=-43 -80
-40
0
40
80
Comprehensive brain coverage of functional activity
already in the first release!

42. Activation similarity fits task similarity
archiemotional
archisocial
archispatial
archistandard
hcpemotion
hcpgambling
hcplanguagehcpmotor
hcprelational
hcpsocial
hcpwmrsvplanguagearchi emotional
archi social
archi spatial
archi standard
hcp emotion
hcp gambling
hcp language
hcp motor
hcp relational
hcp social
hcp wm
rsvp language 0
1
archiemotional
archisocial
archispatial
archistandard
hcpemotion
hcpgambling
hcplanguagehcpmotor
hcprelational
hcpsocial
hcpwmrsvplanguage
archi emotional
archi social
archi spatial
archi standard
hcp emotion
hcp gambling
hcp language
hcp motor
hcp relational
hcp social
hcp wm
rsvp language 0
1

43. IBC reproduces ARCHI and HCP
storyvs.math
social-interactionmotionvs.random
motion
punishementvs.reward
tonguevs.anymotion
rightfootvs.anymotion
leftfootvs.anymotion
righthandvs.anymotion
lefthandvs.anymotion
faceimagevs.shapeoutline
relationalprocessingvs.visualmatching
toolimagevs.anyimage
placeimagevs.anyimage
faceimagevs.anyimage
bodyimagevs.anyimage
2-backvs.0-back
sentencereadingvs.sentencelistening
sentencereadingvs.checkerboard
lefthandvs.righthand
horizontalcheckerboardvs.verticalcheckerboard
mentalsubtractionvs.sentencereading
saccadesvs.fixation
leftorrighthandvs.handpalm
orback
objectgraspingvs.orientationjudgment
social-interactionmotionvs.random
motion
false-beliefstoryvs.mechanisticstory
false-belieftalevs.mechanistictale
expressionintentionvs.genderassessment
facetrustyvs.genderassessment
story vs. math
social-interaction motion vs. random motion
punishement vs. reward
tongue vs. any motion
right foot vs. any motion
left foot vs. any motion
right hand vs. any motion
left hand vs. any motion
face image vs. shape outline
relational processing vs. visual matching
tool image vs. any image
place image vs. any image
face image vs. any image
body image vs. any image
2-back vs. 0-back
sentence reading vs. sentence listening
sentence reading vs. checkerboard
left hand vs. right hand
horizontal checkerboard vs. vertical checkerboard
mental subtraction vs. sentence reading
saccades vs. fixation
left or right hand vs. hand palm or back
object grasping vs. orientation judgment
social-interaction motion vs. random motion
false-belief story vs. mechanistic story
false-belief tale vs. mechanistic tale
expression intention vs. gender assessment
face trusty vs. gender assessment
HCP contrasts ARCHI contrasts
IBCcontrasts
0.6
0.4
0.2
0.0
0.2
0.4
0.6
0.8

44. Variability of Functional Signatures
Group-level z-maps
qFDR < 0.05

45. Variability of Functional Signatures
0.0 0.1 0.2 0.3 0.4
sentence reading vs. sentence listening
sentence reading vs. checkerboard
left hand vs. right hand
horizontal checkerboard vs. vertical checkerboard
mental subtraction vs. sentence reading
saccades vs. fixation
left or right hand vs. hand palm or back
object grasping vs. orientation judgment
social-interaction motion vs. random motion
false-belief story vs. mechanistic story
false-belief tale vs. mechanistic tale
expression intention vs. gender assessment
face trusty vs. gender assessment
face image vs. shape outline
punishement vs. reward
0.0 0.1 0.2 0.3 0.4
tongue vs. any motion
right foot vs. any motion
left foot vs. any motion
right hand vs. any motion
left hand vs. any motion
story vs. math
relational processing vs. visual matching
social-interaction motion vs. random motion
tool image vs. any image
place image vs. any image
face image vs. any image
body image vs. any image
2-back vs. 0-back
read pseudowords vs. consonant strings
read words vs. consonant strings
read words vs. read pseudowords
read sentence vs. read jabberwocky
read sentence vs. read words
Average and standard deviation of map correlation across subjects

46. Effect of subject and task on brain activity
Group-level z-maps qFDR < 0.05
x=10
L R
z=10 -28
-14
0
14
28
L R
y=-50
Subject effect
x=10
L R
z=10 -37
-19
0
19
37
L R
y=-50
Condition effect

47. Effect of subject and task on brain activity
Group-level z-maps qFDR < 0.05
x=10
L R
z=10 -28
-14
0
14
28
L R
y=-50
Subject effect
x=10
L R
z=10 -37
-19
0
19
37
L R
y=-50
Condition effect
IBC data suitable for cognitive and
individual-brain modeling!

48. Future releases
Next release is coming up soon!
Mental-Time Travel
Positive-Incentive Value
Theory-of-Mind + Pain Matrices
Visual Short-Term Memory + Enumeration

49. Future releases
Next release is coming up soon!
Mental-Time Travel
Positive-Incentive Value
Theory-of-Mind + Pain Matrices
Visual Short-Term Memory + Enumeration
Other releases:
Visual system: Passive-Watching of Naturalistic Scenes + Movie
Watching + Retinotopy
Anatomical data + Resting-State
Self-Reference Effect, Tonotopy and more...

50. Work-in-progress
Data acquisition till 2022
Final dataset: 50 acquisitions per participant

51. Encoding models for cognitive mapping

52. Encoding models for cognitive mapping
Dictionary Learning

53. Encoding models for cognitive mapping
Model of occurences of mental concepts

54. Encoding models for cognitive mapping
Model of occurences of mental concepts

55. Thanks!
Bertrand Thirion
...and to
the IBC volunteers!!!