The Basilar Membrane

In the past two weeks, I’ve focused on building the different biological processes in and around the basilar membrane - a key transducer which converts mechanical energy (from the waves generated by the middle ear) into electrical neural signals.

Development Activity - https://github.com/akhil-reddy/beads/graphs/commit-activity
Building the OHC components - https://github.com/akhil-reddy/beads/blob/main/beads/core/cmu/sequencing/combine/audio.py

Please note that some code (class templates, function comments, etc) is AI generated, so that I spend more of my productive time thinking and designing. However, I cross-verify each block of generated code with its corresponding design choice before moving ahead.

Algorithms / Development

Push implementation and transportation

1. Define 200 logarithmic scale tonotopic pitches from 20 Hz to 20 kHz
2. Define the force and pressure mechanics of the membrane

1. Define the tuning mechanisms at each opening
2. Define prestin mechanics for decibel-boosting faint sounds
3. Additionally, prestin also saturates at higher decibels, essentially preventing overstimulation for loud sounds

1. Efferent fibers reduce neurotransmitters onto OHCs to inhibit prestin

Next Steps

Push implementation and transportation

1. Implementing the inner ear CMU components

Building the Environmental Response System (ERS)

1. Building the visual cortex
2. Building the auditory cortex
3. Neurotransmitters - Fed by vision’s bipolar and amacrine cells, for example, to act on contrasting and/or temporal stimulus. Neurotransmitters can be added later during ERU development

Deployment

1. Overlaying video frames onto the retina, including code optimization for channel processing
2. Overlaying audio clips onto the cochlea, including optimization for wave segment processing
3. Parallelization / streaming of cellular events via Flink or equivalent

Why do we need an intricate audio CMU implementation?

Why do we need the inner ear?

Tonotopics is a (refreshing) concept where sound reception is physically organized by pitch (frequency). Here is a simplified illustration - imagine a bunch of sequential holes along a hollow pole, like a flute, where each hole represents a different frequency (100 Hz, 1 kHz, 10 kHz, etc) and is activated only when the instantaneous component of the wave matches its frequency. Collectively, these holes represent the sound wave but the creativity of the breakdown is similar to a fourier transform. This concept is called tonotopics and the hollow pole in this example is the basilar membrane. Biology and evolution in humans has optimized this process even further so that we can hear intricate musical notes as well as loud airplane noises.

Furthermore, a transducer-like biological component is necessary to maintain a consistent neural communication framework - one with spike trains and neural synapses - so that the brain can take advantage of the auditory cortex.

Why do we need the outer ear?

1. Loss of spectral shaping - The outer ear’s resonances (2–5 kHz boost) and head-related filtering are essential for realistic cochlear excitation patterns - skipping them yields a “flat” spectrum at the cochlea, without the boosts and inhibitions at certain frequencies
2. Incorrect level calibration - Without the middle-ear’s ~30 dB gain and compression, we may have to manually estimate how to scale a .wav file’s amplitudes to physiological eardrum displacement. However, errors lead to unrealistic cochlear nonlinearity and neural firing rates
3. Spatial hearing effects - Outer-ear directionality is the basis for localization cues (spectral notches, interaural differences). A direct inner-ear model can’t predict perceived location or spatial unmasking without these nuances embedded into the sound wave