Neural Circuits Flashcards

1
Q

What are 3 questions we need answered to understand how do object recognition?

A
  1. To understand neuronal circuit for single neurons which specifically respond to the presentation of a specific object
  2. Understanding how activation of these neurons causes particular behavioural reactions
  3. Understand if it is one neuron which responds to a stimuli (Jennifer Anniston neuron) or a whole population.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are three factors which we can still recognise the same object despite of?

A

1) Can recognise objects with different orientations

2) Different scales

3) Can recognise objects with a small number of features (E.g. a woman and a man sign outside a bathroom is recognised as a sex despite only made of simple shapes).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the meaning of the hierarchal model of object recognition?

A

At each level of processing, increasing levels of complexity of the stimulus are processed.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What are the 6 levels of object recognition in the Hierarchical model?

A
  1. Detection of edges
  2. Detection of combination of edges and contours
  3. Detection of object parts (e,g, face)
  4. Detection of objects from one point of view (e.g. person or car from the front).
  5. View-invariant object detection (e.g. a particular person or car)
  6. Categorisation (e.g. a human or vehicle).
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What would be the effect of a lesion in the inferior temporal cortex?

A

Decreases ability to recognise objects.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What 2 factors increase throughout the Hierarchical model of object recognition?

A
  1. Increase in complexity of responses of neurons along the ventral stream
  2. Increase in the receptive field size of neurons along the ventral stream

(At beginning, neurons have small receptive field size, but as we move up the hierarchy it gets bigger and bigger)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

How many layers are there in the Lateral geniculate nucleus, what do they receive input from (how are they organised) and of what?

A

> 6 layers organised retinotopically:

  1. First 4 layers receive input from Parvocellular cells
    > Process info on object identity (P ganglion cells recognise objects)
  2. Last 2 layers receive input from Magnocellular cells
    >Processes info on motion (M ganglion cells detect motion)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What type is input from the eyes is used for the layers of the Lateral geniculate nucleus?

A

Each layer receives input from alternative eye (first layer from contralateral eye, second from ipsilateral and so on).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What are 2 ways object recognition is studied?

A

Modelling and electrophysiology to study how neurons map with each other.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What are 2 points a) for b) against the Lateral geniculate nucleus not doing much processing for object recognition?

A

a) It only thalamic “relay station”, ganglion axons make 1:1 connections with LGN projection neurons (info just passed on without processing)
>As receptive fields of LGN neurons are similar to that of ganglion cells it suggests not much processing is done.

b) 6-% of synaptic input to LGN is back propagation from cortex, if info goes backwards to LGN it must do some processing.
>Local interneurons in LGN and these are responsible for processing information (as they change the activity of neurons to process info more efficiently).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What are the 2 visual pathways in the cortex and what do they recognise?

A
  1. Ventral stream/ what pathway
    >Object recognition
  2. Dorsal stream/ where pathway
    >Spatial and Motion recognition
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is the order of the ventral (temporal) pathway?

A
  1. P-ganglion cells
  2. LGN connections with Parvo ganglion cells

> Temporal pathway begins:
3. V1
4.V2
5.V4
6.IT (Inferior Temporal Lobe)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

If the hierarchical model of object recognition is correct, then what is the effect on the ventral and dorsal pathways?

A

The neurons at the start will recognise simple stimuli such as corners at V1 (in ventral pathway) then neurons get more complex such as IT at the end of ventral pathway responds to specific stimuli like faces.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What are the 3 types of columns in the cortex, what do they respond to and what is different between each column?

A
  1. Ocular dominance column
    >Different columns respond preferentially to input from one eye or the other
  2. Orientation columns
    >(respond to direction of edges)
  3. Blobs
    >(respond to colour)

> Columns contain neurons with different properties.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What are the 2 key features of cortex structure?

A
  1. There are 6 layers containing different types of neurons
  2. 3 columns containing neurons with different properties.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What causes ocular dominance columns in the visual cortex?

A

Each column receives inputs from either ipsilateral or contralateral eye

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What inputs onto Blob columns, what pathway are they apart of and what do they recognise?

A

> Receives input from parvocellular cells of the LGN

> Is apart of ventral (temporal) pathway, recognises colour.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What columns are usually found in the V1 cortex and what do they recognise?

A

Orientation columns, these neurons respond to different orientations of edges.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What is a hypercolumn and its function?

A

> Hypercolumn= part of the brain which can encode information about the entire space.

> Made up of the 3 columns of the cortex combined.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

What were 2 results found from Hubel and Wiesel’s experiments using a cat responding to bars of light show?

A

> That in the V1 cortex, neurons in specific orientation columns respond to bars of light in a specific orientation.

> But if the whole receptive field of the simple cells are covered the neurons don’t respond, the bar has to be in the centre of the receptive field of the simples cells.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What layers of the cortex are simple cells localised in?

A

Layers 4 and 6.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What do simple cells in different columns respond to?

A

Simple cells in different columns respond to different orientations.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

What is different about the receptive fields of simple cells (in cortex) compared to cells in the retina and what does this suggest?

A

Simple cells have more elongated and larger receptive fields than cells in the retina. This shows neurons respond to progressively higher complexity of stimulus.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

What did Hubel and Wiesel propose to why simple cells have elongated receptive fields?

A

Neurons which feed to a simple cell are all in one line causing an elongated receptive field.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

What layers of the cortex are Complex cells localised in?

A

Layers 2,3, and 5

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What is the difference in receptive fields of complex and simple cells in the cortex?

A

Simple cells only respond to a bar of stimulus when presented directly in the centre of the receptive field and in a certain orientation, while complex cells respond to a bar anywhere in the receptive field but needing a certain orientation too.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

If a bar of stimulus changes orientation, what is the effect on complex and simple cells responding to it?

A

They would stop responding, as both types of cells only respond to stimuli in certain orientations.

28
Q

Why do complex cells respond to an object with similar orientation in any part of its receptive field?

A

Complex cells collect info from many simple cells with similar orientation, as simple cells respond when objects are in the centre of their receptive fields if all of these are added together then it creates an area where an object with a certain orientation can activate a complex cell.

29
Q

What is suggested by the change in receptive fields between simple and complex cells of the cortex?

A

As simple cells only respond to objects with a certain orientation in the centre of their receptive field, and then complex cells use info from many simple cells to respond anywhere in their receptive field; it suggests the hierarchical model is correct as the neurons’ receptive field increase in complexity.

30
Q

Where are neurons found that respond to complex shapes and and conceive objects such as faces, what pathway is this apart of?

A

Inferior Temporal cortex (temporal lobe), near the end of the Temporal pathway.

31
Q

What does a “grandmother neuron” (Jennifer Anniston neuron) respond to?

A

It is a theory where a specific neurons responds to a particular object (such as recognising your grandma)

32
Q

What is a counter theory to the “grandmother neuron” idea?

A

we perceive a specific object due to a population of specific neurons with similar properties; not just one.

33
Q

What are 3 issues with the Hierarchical model of visual recognition and localisation?

A

1) Does not take into account feedbacks from higher cortical areas.

2) Hard to get experimental evidence as need to record both pre and post synaptic responses in humans

3) We still can’t explain all properties of object recognition.

34
Q

What are 4 examples of object localisation in the visual system?

A

1) Orienting reflex (orientation of the head and eyes to focus important stimulus on the fovea)

2) Smooth pursuit (following moving object)

3) Prediction of motion during prey capture (motion anticipation)

4) Saccadic movements of eyes during object inspection

35
Q

What is the purpose of Saccadic movements of the eye?

A

Entire object can’t fit in fovea so have to move eye so fovea neurons cover all parts of object.

36
Q

Why does a baby not look a auditory stimulus but a child would?

A

As the orientating response involves both auditory and visual systems and this response develops as we grow.

37
Q

Why do we need motion anticipation during stimulus localisation?

A

The processing of visual information takes time (phototransduction takes ~60ms to activate neurons in retina) so we must predict an object’s motion.

38
Q

What area of the brain regulates saccadic movements of the eyes?

A

The superior colliculus

39
Q

What is a Retinotopic map?

A

Organisation of neurons whereby neighbouring cells in the retina feed information to neighbouring places in their target structures (LGN, SC, cortex)

40
Q

What are 3 examples of areas in the brain which contain retinotopic map organisation?

A

Visual Cortex, the Superior Colliculus (SC), and the lateral geniculate nucleus (LGN)

41
Q

If 2 neighbouring neurons were stimulated in the retina, what is the effect on the visual cortex, Superior Colliculus, or Lateral Geniculate Nucleus?

A

If we stimulate 2 neighbouring parts of retina, we stimulate the 2 neighbouring parts of the visual cortex or SC or LGN specific to these 2 retinal neurons

42
Q

In mammals, what part of the brain processes motion and what stream is this found?

A

The MT cortex (middle temporal cortex) is responsible for processing of motion. It is found in the Dorsal stream.

43
Q

What is the role of the dorsal stream?

A

To process motion and spatial recognition to instruct motor neurons to behave in a certain way.

44
Q

What do neurons in the MT cortex respond to and how is this different to V1 simple cells?

A

> Groups of neurons respond to a bar moving in different directions.

> Different to simple cells as they respond to the movement of a bar in any direction as long as they are orientated in a specific way

45
Q

How do neurons in the retina (e.g. ganglion cells) respond to movement?

A

> They are ON/OFF cells so:

  1. Preferred direction= the direction which provokes a response
  2. Null direction= direction of stimulus which causes no response from neurons
46
Q

How is the morphology of a dendritic tree of a neuron involved in processing of motion?

A

The direction of the asymmetric dendritic tree is the preferred direction of the ganglion cell and the cell body side without the dendritic tree is the null direction

47
Q

Why do some Magnocellular ganglion cells have large asymmetric dendritic trees?

A

These ganglion cells are involved in motor detection. So if their dendritic tree is large and asymmetric (only covering one side of cell body) it can respond to motion in one direction while not responding to it in another direction (giving perception of motion).

48
Q

In what area of the retina does direction selectivity arise and what interactions cause this?

A

In the inner plexiform layer, Ganglion cells receive two types of input:

  1. Excitatory input from bipolar cells
  2. Inhibitory from amacrine cells

These interactions allow for ganglion cells to active for one direction of movement, and not be active for the opposite direction (creating directional selectivity).

49
Q

What cell type is directionally selective in the retina?

A

Ganglion cells, specifically
Magnocellular

50
Q

How does input from Bipolar and Amacrine cells mediate direction selectivity in On-Off ganglion cells in the retina?

A
  1. In preferred direction excitation inputs from bipolar cell is larger and Inhibition inputs from amacrine cells is smaller and delayed
  2. In null direction, the excitation is smaller and delayed and inhibition is larger

> Summation of excitatory and inhibitory input causes a larger depolarisation in preferred direction than in the Null direction.

51
Q

What pH is a synaptic vesicle and why?

A

A low pH as many free H+ ions are present inside

52
Q

How does a) depression b) facilitation occur in retinal cells?

A

a) Depression results from depletion of vesicles in the retina

b) Facilitation results from negative feedback from amacrine cells

53
Q

What is the effect of expressing Chanelrhodopsin in the motor cortex of mice and then shining blue light on them?

A

When blue light is on, ChR allows influx of ions (Ca2+, Na+) into neurons firing action potentials, causing mice to run around quickly as motor cortex gets over-excited.

54
Q

How could small organic compounds be used to modify brain activity?

A

> Using molecules which change from Trans to Cis conformation when different wave lengths of light are shining.

> If bound to a protein activator, can make it active when in Trans conformation or deactivate it in Cis conformation

55
Q

Why are small organic compounds used to fix faulty circuits over Channelrhodopsin and Halorhodopsin?

A

As these organic compounds are so small they can fit through pores and into neurons.

56
Q

What is damaged in people with Retinitis Pigmentosa?

A

Photoreceptor layer is damaged and it progresses leading to retinal degeneration.

57
Q

What is the easiest part of the brain to stimulate with electrodes?

A

V1 (visual) cortex

58
Q

If the optic nerve is destroyed, what is the only way to stimulate the visual system?

A

By stimulating the visual cortex directly.

59
Q

Why is it important to activate a faulty circuit early in the visual pathway?

A

As the hierarchical model explains, if we fixed a faulty circuit far up the pathway, then the simple processing needed to see the complex objects would not occur (as this takes place early on in the V1 pathway).

60
Q

What is a disadvantage to activating the V1 cortex with a) Electrical stimulation b) Channelrhodopsin and Halorhodopsin?

A

a) The electrodes warm up when passing electricity, this eventually kills neurons; so cannot stimulate neurons for very long (so not good for sight as is used all the time)

b) Delivering these molecules and expressing in specific neurons is difficult.

61
Q

Why would expressing Halorhodopsin in the degraded photoreceptor layer of someone with retinitis pigmentosa help?

A

Leads to hyperpolarisation of photoreceptor remains in presence of light which mimics what happens in wild type retina. By expressing halorhodopsin we can restore centre surround organisation and directional selectivity of RGCs (retinal ganglion cells).

62
Q

What is needed to control epilepsy?

A

Balance between excitatory release of glutamate and the inhibitory release of GABA

63
Q

How could Halorodophsin and Channelrhodopsin be used to treat epilepsy?

A

Use halorhodopsin in overexcited cells to decrease excitatory input in presence of light, use channelrhodopsin to increase inhibitory output of inhibitory cells in light.

64
Q

What method can be used to get Halorhodopsin or Channelrhodopsin into mice neurons?

A

Optogenetics

65
Q

What are 5 issues with treating epilepsy with small organic molecules, or chnanelrhodopsin and Halorhodopsin?

A
  1. Brain size is too large
  2. Hard to know which neurons to express optogenetics tool (hard to know seizure origin in each individual)
  3. Non-neuronal cells such as astrocytes could be causing the seizures.
  4. Hard to know whether to activate inhibitory or inactivate excitatory neurons.
  5. Hard to detect a seizure before it starts.
66
Q

What does a) IT at end of ventral stream b) MT at end of dorsal stream stand for and what is processed here?

A

a) IT is the inferior temporal lobe where objects are recognised

b) MT is the mid temporal cortex where motion is processed