Brain anatomy and artificial intelligence

Brain anatomy and artificial intelligence

L. Andrew Coward Australian National University, Canberra, ACT 0200, Australia

The Fourth Conference on Artificial General Intelligence

August 2011

Architectures and Information Processes

Call Processing

Registers

Processor

Diagnostics Arithmetic Logic

Control

Maintenance Billing Memory

Functional Architecture

Physical Architecture

Major Brain Structures Lateral Ventricle Thalamus Basal ganglia Hypothalamus Mammillary Body Tegmentum Pons

Precentral Gyrus

Central Sulcus

Post Central Gyrus Cortex Corpus Callosum Fornix Tectum Cerebellum Medulla Spinal Cord

Brain Architecture and Information Processes condition definition and detection

behaviour selection Information flow Thalamus management

Cortex

Hippocampus Resource Management

Behaviour type Amygdala probability management

Behaviour selection

Dorsal Basal Ganglia

Reward management

Ventral Basal Ganglia

Behaviour Cerebellum sequence management

Behaviour Implementation

Brain stem; Spinal cord

Receptive Fields of Cortical Columns

400 µm column All pyramidal neurons in all layers of a column have similar receptive fields

Activity of Cortical Areas During Mental Imaging 9

6

9/46d

46 10

4

8B

3,1,2

33 5

7

8A

9/46v

44 45B 45A 47/12

19

43

32

23

31

7

19

18

11

37

21

38

24

12

25

18 17

34 28

17

22

5

10

39

40

8

9

4

6

3 8

36

35

37

19

20

20 52

42

41

Remembering past event

Conceiving

Elaborating

27

26

29

Imagining future event

30

18

a3 a2

apical dendrite

A B a4 a1

C

soma synapses from different cortical pyramidals with various weights inhibitory synapses from local interneuron

basal dendrite axon

Receptive Fields Must Change as Little and as Rarely as Possible Inputs Cortical layers

Type of condition detected in layer

IV

Conditions that are combinations of inputs from preceding array

II/III

Conditions that are combinations of conditions detected by previous level

V/VI

Conditions that are combinations of conditions detected by previous level

Outputs

Hippocampus gets Information on Internal Column Activity from All Over the Cortex CA1 Hippocampus

Entorhinal cortex

Perirhinal and parahippocampal cortices

Cortex

CA3

DG

Hippocampal Competition Determines Which Cortical Columns will Change Receptive Fields Inputs from entorhinal cortex dentate gyrus

CA3

CA1 Outputs to cortex

pyramidal neuron

inhibitive interneuron

granule cell

mossy cell

Receptive Field Detection in One Area in Response to Different Objects of Same Category BIRD

familiar birds

unfamiliar bird

inactive column

column detecting its receptive field

column expanding and therefore detecting its receptive field

Some Columns are Often (but Not Always) Active When a Bird is Seen BIRD

columns often active when birds are perceived

Columns active when perceiving a novel event

≈ visual objects

≈ groups of objects

≈ group of groups of objects

inactive column

column detecting its receptive field

column expanding and therefore detecting its receptive field

Columns active when hearing words that relate to novel past event ≈ visual objects

≈ groups of objects

≈ group of groups of objects

Capability to indirectly activate columns on the basis of past simultaneous receptive field expansion ≈ visual objects

≈ groups of objects

≈ group of groups of objects

Column population indirectly activated on the basis of past simultaneous receptive field expansion ~ visual objects

~ groups of objects

~ group of groups of objects

= group of groups during original experience

condition definition and detection

behaviour selection Information flow Thalamus management

Cortex

Hippocampus Resource Management

Behaviour type Amygdala probability management

Behaviour selection

Dorsal Basal Ganglia

Reward management

Ventral Basal Ganglia

Behaviour Cerebellum sequence management

Behaviour Implementation

Brain stem; Spinal cord

Behaviour Selection, Including At Least and Only One Behaviour

Cortex

Striatum

D1

D2

Excitatory (glutamatergic)

Dorsal Basal ganglia

Inhibitory (GABAergic) GPe

Modulatory (dopaminergic)

STN

SNc

Modulation pathway

GPi/SNr

Direct pathway

Indirect pathway

Thalamus

Implementation of Strategic, Tactical and Detailed Reward Behaviours VTA

St

}

Shell

r ia do

}

pa m in en eu ro ns

} Dorsolateral striatum

orbital and dorsolateral medial prefrontal prefrontal

premotor

SNc

motor

Cortex

m

a in

} Central striatum

tu

br

} Core

r ia

id

Shell

St

M

m

}

tu

}

Core

Central striatum

}

Dorsolateral striatum

condition definition and detection

behaviour selection Information flow Thalamus management

Cortex

Hippocampus Resource Management

Behaviour type Amygdala probability management

Behaviour selection

Dorsal Basal Ganglia

Reward management

Ventral Basal Ganglia

Behaviour Cerebellum sequence management

Behaviour Implementation

Brain stem; Spinal cord

supplementary motor cortex

hippocampus front

Thalamic reticular nucleus

right

motor cortex; supplementary motor cortex touch, propioceptic sensory information

AN

VA

entorhinal cortex

VL

Massa intermedia

LD MD

VP

somatosensory cortex

LP

P

visual and other association areas

LGN medial prefrontal cortex

prefrontal cortex

primary visual cortex

visual information

MGN

primary auditory cortex

auditory information

Implementation of Release Behaviours by Imposing Frequency Modulation

cortical layers

IV V

cortical pyramidal neuron

VI

thalamocortical projection neuron thalamic reticular nucleus

TRN inhibitory interneuron excitatory connection

thalamic nucleus

inhibitory connection inhibitory connection

basal ganglia

condition definition and detection

behaviour selection Information flow Thalamus management

Cortex

Hippocampus Resource Management

Behaviour type Amygdala probability management

Behaviour selection

Dorsal Basal Ganglia

Reward management

Ventral Basal Ganglia

Behaviour Cerebellum sequence management

Behaviour Implementation

Brain stem; Spinal cord

Parallel Cerebellar Path for Implementing Frequently Used Sequences of Actions

To design an artificial general intelligence system Focus on the information processes that are required If a system needs to

condition definition and detection

behaviour selection Information flow management

Perform many different behaviours Detect many different conditions Define most of the conditions heuristically

Behaviour selection

Limit the resources required

Then the information processes will need to be Condition definition and detection Condition resource management Information flow management Behaviour selection Reward behaviour selection Behaviour priority management Behaviour sequence management

Resource Management

Reward management

Behaviour type probability management

Behaviour sequence management

Behaviour Implementation

a. Boundary of object in visual field

b. Some of boundary element receptive fields detected in V1

c. Retinal area for which one boundary element receptive field detection recommends modulation of inputs from area

d. Total recommendation strengths for modulation much stronger for area within object boundary

postsynaptic potential

threshold

total

5 input action potential spikes

10

15

20

25

30

35 msec

source neurons

target neuron

integration window close to modulation minimum

integration window close to modulation maximum

time window of one third of the modulation cycle

I

II

III

entorhinal cortex

dentate gyrus

amygdala supramammillary area

II subicular complex III CA3

mammillary bodies

V/VI

CA2

anterior thalamus

CA1

cortex

pyramidal neuron granule cell inhibitive interneuron mossy cell

Purkinje cell dendritic tree perpendicular to parallel fibres

Connectivity

Purkinje cell bodies parallel fibre outputs from granule cells

granule cells Purkinje cell output targets cerebellum nuclei Pontine Nucleus

mossy fibre input targets granule cells

climbing fibre input targets small group of Inferior Olive Purkinje cells (~10)

9

6

9/46d

33

3,1,2 5

44 45A 45B

40

47/12

19

42

12

18 17

34 28

25

20 52

11

37

21

38

7

19

18 17

22

23

31

10

39

43

5

24

32

9

4

6

8

7

8A

9/46v

46 10

4

8B

3 8

36

35

37

19

20

41 27

Areas concerned with working memory

Areas with receptive fields corresponding with groups of cortical columns often active at similar times in the past Areas with receptive fields corresponding with groups of cortical columns that expanded their receptive fields at similar times in the past

26

29

30

18

striatum

left cerebral basal nuclei cerebellum hemisphere

lateral ventricles

some of basal ganglia

thalamus

i

ii

iii

vi

v

iv

caudate putamen substantia nucleus nigra nucleus accumbens

fornix hippocampus septal nuclei

hypothalamus

claustrum

amygdala

the striking commonalities in medial left prefrontal and parietal activity during the elaboration of (a) past and (b) future events (relative to the control tasks) From Addis, D.A., Wong, A.T. and Schacter, D.L. (2007). Remembering the past and imagining the future: Common and distinct neural substrates during event construction and elaboration. Neuropsychologia 45, 1363-1377.