Notesale: Turn your study into money

Document Preview

Extracts from the notes are below, to see the PDF you'll receive please use the links above

---

11
...
Use the
time difference of when the sound arrives at each ear- inter-aural time difference
...
This is not
an easy problem to solve
...

So how would you shift spike trains from left ear and spike trains from right ear to cause them to match to
identify the sounds- by cross correlation
...
Integrate the probability of getting spike from A at certain time (t) multiplied the
probability of getting a spike from B at that same time plus delay (t +T) integrated graded the entire time of
the recording
...


PAB (Τ) = ∫ PA( t )PB( t+Τ ) dt

Jeffress though it might be possible to have slow unmyelinated axons carry signals from the different ears
using different lengths of axons
...
If that travel time difference exactly
matches the actual inter-aural time difference, signals will arrive simultaneously from left and right ears
...
The
frequency of firing is proportional to product of inputs to ensure coincidence detection:
O= XL XR, if XL= the probability of stimulus onset in left ear
...
12
Leaky integrate and fire neuron model does coincidence detection
...
e
...

If time constant T << period (1/frequency) of inputs from L or R ear and threshold is near 2, then model
neuron only fires for coincident inputs, and effectively multiplies the probabilities of getting a spike on either
input

Pout( spike in [t, t+T] ) = PL ( spike in [t, t+T] )x PR ( spike in [t, t+T] )
Firing rate = Pout( spike in [t, t+T] ) / T
Problem:
Neurons respond to different pitches of sound
...
The inputs
are continuous, so at the first time interval there is a peak showing a match of sound from right and left ear
...


NL: nucleus laminaris have cross correlation array of neurons which are detecting time differences, and they
are arranged tonotopically
...
Here, the neurons sum ITD across different frequencies, so only the neuron with the correct
inter-aural time difference has a high firing rate
...
Finally in the ICx, there is another row of neurons corresponding LS neurons
...
This introduces winner takes all architecture to produce clean peaks
...
Auditory cues are also
aligned in retinal co-ordinates, so that both visual and auditory stimuli will generate eye movement to the
direction of the stimuli
...
This
arrangement is simpler in barn owls because they move their heads rather their eyes, primates can move
their eyes
...
In barn owls one ear points a bit lower and one ear points a bit more
upwards
...
However this is not the case in lots of
other animals so this is a lot more complicated
...
12
Colliculus generating an eye movement is a nice example of population vectors
...
Quite a long time ago, people working on eye
movement realised that the inferior colliculus generate eye movement in a population average sort of way
...
So the population vector, or the average firing rate weighted average vector location will be where you
move your eye to
...
1976)
...
When animal tried to move to A, the movement was normal because the average sum of the
neurons that was remaining still had a population vector towards A
...
Movement to C is deflected downwards
...

Temporal processing in olfaction
Odour masking problem (Ambrose-Ingerson et al
...
g
...
They have a sniff cycle which is 3
to 5 sniffs at 10 Hz and whiskers move at the same frequency
...

In the olfactory bulb: Triangles are mitral cells (excitatory), circles are granule cells (inhibitory)
...
Total amount of activity is normalised, whether it is a weak
or strong smell you tend to get the total amount of activity which implies there is feedback inhibition
...
There is
layer II pyramidal cells (excitatory) and various interneurons
...
Then layer II pyramidal cells in the piriform cortex projects back to olfactory
bulb onto the inhibitory granule cells
...
So they thought the excitatory connections
between the layer II cells in the piriform cortex could form an associative memory or a positive feedback

11
...
So neurons which are always representing the same familiar odour, tend to fire at the same time
because they both get inputs from mitral cell
...
So if
there is a little bit smell of chocolate, it will activate other neurons that respond to chocolate
...
Different subnets represent different smells
...
The layer II neurons always project back inhibitorily into
the olfactory bulb
...
This is a neural implementation of competitive
queuing
...

The authors researched this mechanism via a
simulation experiment
...
A was twice as strong as C
...
After learning of
what A and C smells like separately, on the first
sniff of A/C odour mixture, all the A neurons
reactivate each other and inhibited firing of C
neurons
...

Then the second sniff C neurons were active as
neuron A have inhibited each other

---