Hearing (pt2): Frequency coding, intensity coding

Frequency coding in cochlea (Active passive wave)

• Active+passive process wave: peak is much narrower & higher in living ear (So active + passive is more efficient in frequency coding) than just passive

  • Active: Involve in brain filtering of sound info

  • Passive: just physical stimulation of basilar membrane

• Passive process wave: based on physical properties of basilar membrane

• Smth is sharpening & amplifying the response – cochlear amplifier

Otoacoustic emissions

• Movement of outer hair cells creates sounds that exit the ears (use these emissions to diagnose impaired vs healthy hearing)

• Byproduct of electromotility

• Evoked emissions vs Spontaneous emissions

• Reduced by aspirin (reduces activity of outer but not inner hair cells)

• Taken as evidence for outer hair cell involvement in active process/ cochlear amplifier

Evoked emissions

occur in response to auditory stimulation

• depend on frequency of stimulating sound; now used clinically as quick indicator of inner ear damage

Spontaneous emissions

occur without stimulation (less than 20 dB, 1000-2000 Hz

Active process (cochlear amplifier) — Electromotility

• Sharpens & amplifies frequency response of basilar membrane relative to frequency response based on passive, physical properties of the basilar membrane

• Due to the electromotility of outer hair cells…

  • Depolarization causes contraction (shortenting); hyperpolarization causes elongation (lengthening) at frequency of incoming sound wave

  • This motion (contract/lengthen) is powered by a special motor protein called Voltage-sensitive protein (prestin)

  • Otoacoustic emissions are a byproduct of electromotility

  • Makes otoacoustic emissions

What does electromobility do exactly?

It’s an cochlear amplifier!

Amplifies basilar membrane motion: adds extra mechanical energy back into the cochlea, making the vibration stronger and sharper

Improves frequency selectivity: Makes nearby inner hair cells more sensitive to faint sounds.

Helps detect quiet sounds: Tuning curves sharper, distinguish small differences in pitch

Function of inner hair cells

• convey most sensory info about sound to brain

  • Receptor cells like visual rods and cones

Function of outer hair cells

• Modulate sensitivity and sharp frequency-tuning of cochlear partition

  • Brain sends efferent electrical potentials to cochlear

  • When they are stimulated, they lengthen and extend farther into the tectorial membrane

  • Through lengthening and contracting (electromotility), outer cells cause parts of the cochlear partition to stiffen in ways that make the responses of inner hair cells more sensitive, and more sharply tuned to specific frequencies

Frequency coding in auditory nerve (also called place coding)

• Absolute threshold for individual auditory nerve fibres as a function of frequency (threshold tuning curves)

• Frequency selectivity is clearest when sounds are very faint

• Lowest threshold at CF (Low threshold, high intensity)

• Cells close to helicotrema (away from oval window) respond to lower frequencies; cells close to oval window respond to higher frequencies (at least for faint sounds)

How does neuron respond when the sound is faint (low intensity)?

When sounds are faint, the neuron responds only to its best frequency → sharp frequency selectivity.

When sounds are loud, the neuron responds to a wider range of frequencies → less precise selectivity.

Characteristic frequency (CF)

Frequency that increases the neuron’s firing rate at the lowest intensity (dB) (lowest y-axis point on the threshold tuning curve) (The y-axis is the threshold intensity (dB) required to stimulate fiber above spontaneous firing rate)

• A low-intensity sine wave tone w a certain frequency will cause certain AN fibers to increase their firing rates, while other AN fibers continue to fire at their spontaneous rates

• Dash line shows what tuning function for Fiber 6 would look like if outer hair cells are absent

  • Presence of outer hair cells improves the sensitivity (threshold intensity for firing above spontaneous rate) and the frequency selectivity (sharper tuning curve) of auditory nerve fibres

Two-tone suppression

Decrease in firing rate of auditory nerve fibre to its CF when a 2nd tone of similar frequency (shaded areas only) is presented at the same time

• Occurs in the outer hair cells

• Frequency code for complex sounds is NOT the sum of indv auditory nerve fibre responses to indv pure tones

• When the ear is stimulated with two tones at the same time—say, a primary tone (the one you’re testing) and a second nearby tone (the suppressor)—the response to the primary tone becomes smaller than it would be if that tone were presented alone. In other words, the second tone suppresses the cochlear response to the first.

When is the suppression effect esp stronger?

if the second tone (suppressor) tone has a lower frequency than the first tone

E.g. If you play a 1,000 Hz tone at a moderate level, the cochlear response is strong. But if you add a second tone at 1,100 Hz, the response to the 1,000 Hz tone diminishes. (But second tone suppresses first tone more if it’s 990 Hz)

“Decrease in firing rate of auditory nerve fibre to its CF when a 2nd tone of similar frequency is presented at the same time.” What’s the function?

• Each auditory nerve fibre has a characteristic frequency (CF) — the frequency it responds to most strongly.

• If you present a tone at the CF, the fibre fires vigorously.

• But if you present a second tone that’s close in frequency (and often at a slightly higher intensity), the firing rate to the CF tone decreases.

Isointensity curves  for single neuron (each color is at diff intensity)

• The bottom curve (20dB) is avg firing rate (action potentials per sec)

• Neurons become less sharply tuned w increasing sound lvl (dB). 

• At low intensities (e.g. 20 dB) → curves (neurons) are sharply tuned, peaking at the characteristic frequency (CF) (The fiber only responds to frequencies near its CF)

• At higher intensities (e.g. 60 dB) → tuning broadens, so the fibre responds to a wider range of frequencies.

• Because of rate saturation: once the fibre maxes out at its CF, it can’t encode increases anymore, but off-CF inputs still drive activity, so the fibre appears less frequency-specific.

Why do neurons become less sharply tuned at higher intensity?

At low intensity (The fiber only responds to frequencies near its CF)

• Basilar membrane vibrations are small → Outer hair cells amplify motion most effectively right at the CF location

At higher intensity

The sound wave causes larger basilar membrane displacements (wave travels larger and spreads farther)

Stereocilia of nearby hair cells (away from the exact CF place) also get displaced enough to trigger firing.

Auditory nerve fibres that were once “quiet” at those off-CF frequencies now fire more.

Rate saturation

Point at which an auditory nerve fibre’s firing rate no longer increases with increasing stimulus intensity.

• Each AN fibre has a maximum firing rate.

• Once stereocilia are already pivoting as far as they can, increasing intensity (or slightly changing frequency) doesn’t make the neuron fire faster.

  • Example: At 1500 Hz, the fibre saturates, so both 1500 Hz and 2000 Hz tones produce maximum firing at high intensities.

• Place coding can’t be the whole story for sounds of higher intensity

Population coding as the solution cuz Relying on one fibre is unreliable (cuz rate saturation – each AN fiber has CF and max firing rate)

Auditory nerve fibres from diff regions of the basilar membrane fire when intensity is constant but frequency is changed

The auditory system solves this by pooling information across fibres with different spontaneous rates: High-spontaneous fibres and Low-spontaneous fibres

High-spontaneous fibres

Fire a lot even at low intensities (in silence), but saturate quickly (hit max firing rate quicker than low-spon)

Low threshold: Respond to very quiet sound

Maximum rate below 60 dB

Thicker axon

Low-spontaneous fibres

Fire slowly at low intensities (in silence), need higher intensity to respond, but don’t saturate as quickly.

High threshold — need louder sounds to respond

Maximum rate above 60 dB (Thinner axon)

Spontaneous activity

Even with no sound, auditory nerve fibres fire at a baseline (resting) rate of action potentials

High spontaneous fire well at no sound, low spontaneous still fire, but slow rate and saturate much slower

Stereocilia

Tiny hair-like projections on top of hair cells in the cochlea.

• They’re mechanically sensitive: their deflection (bending) opens or closes ion channels

Resting and stimulated firing

• AN fibers fire at a baseline (resting) rate even with no sound.

• When a tone matches the fibre’s characteristic frequency (CF), stereocilia displacement is strongest → firing rate rises well above baseline

In addition, place of maximum firing codes frequencies > 500 Hz, but

• At lower frequencies, basilar membrane envelope is too broad for good frequency discrimination

• But, frequency discrimination is good below 500Hz

• Frequency of firing provides a temporal code for frequencies below 1000 Hz (e.g. 100 spikes/s fired to 100 Hz tone)

Phase locking

Auditory nerve fibre fires at 1 distinct point (phase) in cycle of sound wave

• Provides temporal code for sound wave frequency

• The histogram (blue, bottom) shows neural spikes for an auditory nerve fiber in response to the same low-frequency sine wave (red line, top) being played many times. 

  • Note that the neuron is most likely to fire at one particular phase of each cycle of the sine wave

Temporal code

A coding strategy where the timing of auditory nerve spikes carries information about the frequency of a sound. 

• When a pure tone causes the basilar membrane to vibrate, the auditory nerve fibres tend to fire in sync with a particular phase of the sound wave (phase locking)

At what frequencies that fibres can’t fire on every cycle? (Volley principle is solution)

1000 Hz. cuz Max firing rate of an auditory nerve fibre ≈ 1000 spikes/sec.

Combined firing of a group of fibres matches frequency of incoming sound (1000-4000 Hz) to provide a temporal code for frequency

• Problem: A single neuron can’t fire fast enough to keep up with high-frequency sounds.

  • But humans hear up to 20,000 Hz!

• Solution: Fibres “team up.”

  • Different neurons fire on different cycles of the sound waveform.

    Individually, a single neuron may skip cycles because of its firing limit.

  • Population of neurons fire collectively in a pattern

  • The combined firing of the group preserves the timing of each cycle (preserve temporal code — to keep spiking every cycle carries info to brain)

  • This allows temporal coding to work for frequencies well above the firing rate limit of any single fibre (up to ~4000–5000 Hz in humans)

How can firing rate be the code for frequency and for intensity?

Frequency: Temporal coding (timing of spikes)

• Phase-locking; volley principle; spike timing matches waveform cycles

Intensity: Rate coding (firing rate of spikes)

• Higher sound intensity → vibrate more of basilar membrane → depolarize more → higher spike rate and recruitment of more fibers