L07 & L10- Light and Dark Adaptation Flashcards
Mechanisms of retinal adaptation during:
1) Sunrise to sunset (global, slow, large)
2) Eye movements (local, fast, small)
1) Changes in pupil size, switch from cone to rod vision, changes in photopigment content of photoreceptors
2) Changes in the gain of phototransduction and postreceptoral circuitry
In terms of vision: what is adaptation? Weak inputs are lost in ___. Strong inputs saturate the __ response
Adaptation acts to match the limited neuronal response range to the visual input range
Noise, GC
*GCs have a response range similar to a digital camera
Long exposure makes things darker/brighter?
Brighter - more detail in shadowed region
Does retina use local or global adaptation?
Local
Laughlin’s Rule - says nothing about spatial extent or dynamics
Encoding efficiency is maximized by a stimulus-response relation that makes each response equally likely given the distribution of input signals
Natural images contain info in the same or many different statistical moments of the intensity distribution? What two factors are most important?
Different Mean intensity (intensity overall) and contrast (differences in intensity in a scene)
Is fixation stable?
No, it is small, random movements.
Behavioural threshold is determined by what? What is the problem with this?
Combination of the gain of the phys response to light and the noise that obscures this signal - PROBLEM: can’t tell whether an increase in threshold is due to a decrease in gain and/or an increase in noise
“Dark light” region (rods and cones) - Is threshold dependent or independent of background? What is the dark light?
Independent
Dark light refers to spontaneous activation of rhodopsin that generates intrinsic noise
Rose-DeVries region (rods)
a) Threshold increases in proportion to?
b) Classical explanation? Increase in background intensity = ?
c) Support for classical explanation?
d) Alternate explanation? Increase in background intensity = ?
e) Support for alternate explanation
*Explanations are not mutually exclusive
a) Threshold increases in proportion to the square root of the background
b) NO adaptation - photon noise obscures the test stimulus. Background lights produce noise (due to statistical fluctuations in the number of quanta delivered by a fixed intensity background)
We cannot control how much quanta we get - there is variation. (Poisson distribution is the probability distribution that governs how many quanta is delivered. The variant is the square root of a mean). Increase in background intensity = increase in noise = increase in flash to compensate
c) Parsimonious model - simple explanation based on physics
d) Adaptation - photon noise is regulated by adaptation. Increase in background intensity = decrease in gain = increase in flash to compensate
e) Suprathreshold brightness matching -situation where noise is irrelevant leads to a slope of a 1/2
Toad RGC show square root adaptation but mammalian RGCs do not
What is gain?
It is a type of adaptation that controls level of response
High gain - amplifies response from neuron
Low gain - decreases response from neuron
e.g. High input going into system means you will use a low gain and vice versa
Weber regions (cones)
a) Threshold increases in proportion to?
b) It has a slope of _ (what number)
c) Is contrast independent or dependent on the intensity of the illuminating light?
d) Explanation
e) The explanations are based from behavioural measures, what factors do they not take into consideration?
a) Threshold increases in direct proportion to background
b) 1 - as threshold and background are directly equal, if threshold = 1, background = 1, then slope is 1/1 = 1
c) Contrast in an image is independent of the intensity of the illuminating light
e. g. Car on the road illuminated by sunlight. Decrease in sunlight affects both the car AND road - hence ratio doesn’t change and contrast is the same. Even if you wear sunglasses during the day - it doesn’t affect what you can or can’t see as you cut down on both the amount of light coming from the stimulus (e.g.car) and light by the same proportion
d) Cone photoreceptors and GCs show Weberian adaptation
e) The fact that light adaptation causes changes in response kinetics
- e.g. Faster responses in cones, horizontal cells and RGCs
- These changes are not apparent in the TVI data
- Role of noise is not well understood
Weber’s law
The law states that the change in a stimulus that will be just noticeable is a constant ratio of the original stimulus
What happens to Weber’s law when you wear sunglasses at NIGHT?
Uses your rods instead where the law is non-existent and you won’t be able to see, low contrast things = invisible, contrast is no longer constant
There is always uncertainty. Low light levels = noisy from stimulus noise or spontaneous firing, hence adaption will be noisy. How do you deal with this?
Spatial and/or temporal pooling reduce stimulus and photoreceptor noise
1) Pool photoR then average = reliable response
2) Responses increase at subsequent stages of retinal processing
3) Adaptation at late stages - reduces risk of fluctuations in gain, but increases risk of saturation