Psychophysics
Sensory Neurons and Receptor Potentials
Receptor potentials are graded potentials generated by sensory transduction, which converts physical energy into electrical signals. This physical energy can be:
Electromagnetic radiation (photoreceptors)
Mechanical force
Chemosensory stimuli
Receptor potentials are graded, meaning that the stimulus amplitude is coded by graded changes in membrane potential.
Example: In vertebrate retinas, light causes hyperpolarization. The hyperpolarization is graded with the intensity of the light flash. Superimposed traces are centered around the time of the light flash, with each trace representing a response to a flash of increasing intensity, eventually plateauing.
More commonly, receptor potentials are depolarizing, as seen in invertebrate photoreceptors. These responses to light are also graded in amplitude from dimmest to brightest flashes.
Receptive potentials are generated by the selective opening or closing of ion channels. Depolarizing receptive potentials occur due to the opening of ion channels.
Limits on Receptive Potential Amplitude
The amount of change in a receptive potential is limited, which limits the range over which a graded receptive potential can code stimulus intensity. To understand this, consider the ionic basis of the receptive potential.
An excitatory postsynaptic potential (EPSP) has a reversal potential typically near zero millivolts. This doesn't correspond to a single ionic reversal potential. In an EPSP, ligand-gated cation channels allow passage of both sodium and potassium.
Potassium flows out, and sodium flows in.
The potassium current carries the membrane potential towards the potassium equilibrium potential.
The sodium current drives the potential towards the sodium equilibrium potential.
The result is a compromise reversal potential near zero millivolts, lying between the sodium and potassium reversal potentials.
The situation is similar for many depolarizing receptive potentials. A generalized cation increase leads to a reversal potential between the sodium and potassium reversal potentials, around zero millivolts.
At rest, transduction channels are closed. A small stimulus opens a proportion of the channels, resulting in an increase in conductance and a depolarization proportional to the increase in conductance. The stronger the stimulus, the more channels that open.
At some point, the number of open channels drives the membrane potential close to the reversal potential. Increasing the proportion of open channels further will not cause the membrane potential to keep depolarizing; it can only approach the reversal potential. As the membrane potential gets closer to the reversal potential, further increases in conductance cause ever smaller increments in membrane potential change. Opening the remaining channels will not significantly change the membrane potential because the open channels drive the membrane potential towards the reversal potential.
In sensory transduction, stimulus intensity is coded by arranging for increasing stimulus strength to open more transduction channels. However, this only works within the range between the resting potential and the reversal potential.
Ideally, close to 100% of channels would be open with stimuli that bring the membrane potential very close to the reversal potential, and a small percentage would be open for small stimuli. Over this range, graded changes in membrane potential are proportional to stimulus intensity, and these changes are proportional to the increase in conductance caused by the increased number of open transduction channels.
For onward transmission, these graded potentials are translated into a spike frequency or spike rate. A small stimulus generates a low frequency of action potentials, a medium stimulus generates a medium frequency, and a maximal stimulus generates a maximal frequency of action potentials.
These ionic mechanisms impose limits on sensory coding.
Logarithmic Sensitivity
Given the limited range of receptive potentials and the wide range of natural stimuli, receptors can be logarithmically sensitive.
The resting membrane potential is set to zero to read off the amplitude of the receptive potential, where the y-axis represents the voltage change from rest ([]).
Receptors encode stimulus intensity logarithmically, allowing them to respond to a wide range of stimulus intensities by compressing the stimulus range into a manageable neural response range.
A plot of peak amplitude of voltage increase from rest versus relative flash intensity shows that intensities are closely spaced initially but spread out at higher intensities.
When the x-axis (relative flash intensity) is converted to a logarithmic scale, there's a reasonably linear relation, where response amplitude is proportional to the logarithm of the intensity. Then, the response begins to saturate.
This logarithmic coding is typical. The curve derived from plotting this data includes:
A threshold - the point at which a reliable response above noise can be detected.
A region of graded linear response.
A saturation point
Between the threshold and saturation points lies the dynamic range, which is the range over which the system can effectively cope.
There is a trade-off between sensitivity and dynamic range:
A large dynamic range, which allows detection of both weak and strong stimuli, requires a curve with a more gentle slope.
High sensitivity to differences requires a steeper slope.
If the response magnitude is plotted against the stimulus intensity:
A relatively small increase in intensity will produce a small increase in response with a gentle slope.
The same small difference will translate into a much larger difference along the y-axis with a steeper slope, making small intensity differences easily resolvable.
However, a system with a steeper slope will saturate more quickly, reducing the dynamic range. Conversely, a system with a gentle slope covers a larger range of intensities but has lower sensitivity to differences.
Solutions to this trade-off include:
Adaptation: Adapting to one level of stimulus intensity means shifting the curve along the x-axis to operate within different intensity levels.
Range fractionation: Using different receptors with the same sensitivity to differences but over different absolute intensity ranges. For example:
One population of sensory receptors has a very low threshold.
Another has a somewhat higher threshold but is tuned similarly to differences over a higher stimulus intensity.
An example of range fractionation is the duplex retina: rods and cones.
Rods are sensitive to low light levels but saturate at moderate intensities.
Cones are not stimulated at all in very low light intensities but are operative over higher light intensities.
Psychophysics
Psychophysics measures sensations and perceptions and correlates them with neural activity. It has a history of well over a century and follows discoveries in physics.
Measuring Sensations
Physical stimuli, such as light, can be measured as electromagnetic radiation. Associated sensations, such as brightness or sensitivity to contrast, can be measured. Stimulus intensity can be expressed in power units, such as watts per unit area; in practice, more convenient measures are often used.
The subjective sensation of brightness cannot be measured directly, but we can:
Measure thresholds to determine the limits of sensitivity
Use scaling to order stimuli in perceived intensity along a perceptual dimension
A threshold can mean two different things:
Absolute sensitivity: the dimmest light that can be perceived or the quietest sound that can be detected.
Sensitivity to differences (difference threshold): the smallest difference that can be detected between two stimuli.
Absolute Threshold vs. Difference Threshold
A difference threshold is also called the just noticeable difference (JND). To measure an absolute threshold, a tone is played at varying volumes, and the subject answers yes or no regarding whether they heard it. When the tone is very low intensity, below human hearing, the response is consistently