Journal Club: Finite element modelling of the cochlear motion - with realistic Osseous Spiral Lamina structure based on a human cochlea

on

Today's journal article

A, Lennon JB, Maini R, Tan X, Niazi A, Park J, Park S, Richter CP, Ebeid M. Kölliker's Organ Functions as a Developmental Hub in Mouse Cochlea Regulating Spiral Limbus and Tectorial Membrane Development. 

Why I picked this article

The hearing organ, the cochlea, is a super-sensitive biological sensor of sounds. The spiral shape of the cochlea, along with other anatomical features, is designed to detect extremely small changes in pressure coming from sounds. Our current understanding of how the cochlea does this largely comes from small animal models like gerbils and mice. However, the human cochlea isn’t just a scaled-up version of other mammals. 

The space inside the cochlea is divided into cochlear compartments by bony and soft structures. One of the bony divisions that runs along the cochlea is called the osseous spiral lamina (OSL). The OSL bulges out from the central bony part of the cochlea and runs towards the edge of the spiral, and provides a bony division within the cochlea. Within the OSL, there are two thin bony plates (vestibular plate and tympanic plate) with a space between them that carries auditory nerve fibres. The edge of the OSL connects to a soft-tissue bridge to the basilar membrane. Interestingly, the soft-tissue bridge is a human-specific feature not observed in small rodent models. 

Many computerised models consider OSL to be a very stiff or rigid plate. This research focuses on building a human-specific finite element (FE) model that explicitly includes these features of OSL and checks it against motion measurements from cadaver temporal bones. It helps explain how the human cochlear partition actually moves in response to high-frequency sounds.

Some of the research findings

Finite Element Modelling (geometry): 
  • Tapered straight, rectangular cochlear model (4 mm x 6 mm at the base and 4 mm x 0.6 mm at the apical end)
    • Two fluid chambers (4 mm x 3mm - 4 mm x 0.45mm, and 4 mm x 3 mm - 4 mm x 0.15 mm) corresponding to scala tympani and scala vestibuli, respectively. 
    • A cochlear partition, oval window (2.8 mm x 1.3 mm) and round window (1.2 mm x 1.6 mm)
    • Cochlear partition includes a wide, two-plate OSL and a soft-tissue bridge to the basilar membrane. 
  • OSL: 1.1 mm (base) and 0.3 mm (apex)
    • Plates within the OSL: vestibular plate height and tympanic plate height are both 9 μm, the middle layer is 22 μm
    • OSL-basilar membrane bridge is 0.2 mm (base) and 1.0 mm (apex) 
Finite Element Modelling (Material properties) 
  • Based on prior literature and adjusted to match a passive human frequency–place map as summarised in Table 1 within the article. 
  • e.g. OSL bone: Poisson's ratio 0.3, Density 2,000 kg/m3, Young's modulus 10,000 MPa, Damping 0.1
  • e.g. Round window membrane: Poisson's ratio 0.485, Density 1,100 kg/m3, Young's modulus 1.4 MPa, Damping 0.8
  • Experimental target: CP velocity normalised to stapes velocity from seven human cadaver specimens with best frequencies 9.5–14.4 kHz. The model was tuned and then compared against these measurements in magnitude and phase.
Data Availability Data and Finite Element model (in COMSOL ver 6.1) for this study are available on the Harvard Dataverse site https://doi. org/1 0. 7910/ DVN/ NTUHYI 

Figure 2. Finite element modelling - geometry summary. Tubelli et al. 2025

Key findings: 
  • Motion match: Basal CP motion from the model followed the experimental trends in both magnitude and phase in the high-frequency region.
    • When compared to the cadaveric data, the variability between each bone was very high. 
    • Nonetheless, the motion along the cochlea (along different frequencies) was matched with the OCT data. 
  • When radial cross-section (from OSL to the basilar membrane) was considered, the model mimicked the peak magnitude of movement at the junction point between the OSL-bridge and the start of the organ of Corti. This pattern was maximised at the best frequency and diminished as the frequency was further out from the best frequency. 
  • The motion and phase of different layers were very similar (very small deviation of each component) 
  • OSL stress on OSL: Modelling showed that the stress on the OSL was concentrated at the two plates within the OSL, near the bridge at the best frequency, and scattered more towards the modiolus as the frequency decreased from the best frequency. 
  • OSL stress (inside): Modelling showed negligible stress along a middle layer of OSL, containing the neutral axis, compared with the outer OSL plates. 
    • This suggests that the three-layer OSL’s neutral axis may provide a relatively stress-free passage for auditory nerve fibres.

Haruna's takeaway

This is a very nice modelling publication - not only is the modelling valuable, but also the presentation schematics were very helpful for a non-modeller like me to still be able to read the publication. The osseous spiral lamina (OSL) is such an interesting structure, as we see it being much thicker and bony in sheep, which has a similar (~70-80%) size to the human cochlea. The very fragile nature of the human OSL must have some implications for the cross-species difference in hearing. I do wonder if the thinner human OSL is more vibratory than sheep OSL, or if it's simply a reflection of more nerve fibres needed to travel within the OSL? It would be great if we could change the OSL plate thickness in the model. 

 ------- 

This is Haruna's 83/100 of the 100-day challenge to post a science blog article every day! I love inner ear biology & cochlear physiology.