Semester project - Modeling the Human Vestibular System Encoding Using a Neuromorphic Chip

This project aims to model the human vestibular system's afferents using the Dynap-SE1 neuromorphic chip. The vestibular system, essential for balance and spatial orientation, comprises the otolith organs and semicircular canals,which encode head position and velocity relative to space [1, 2]. Otolith organs detect linear acceleration, while semicircular canals sense angular velocity. Vestibular afferents are classified as regular or irregular based on their firing patterns and morphology [3]. Regular afferents employ temporal coding for precise stimulus timing, whereas irregular afferents, with their higher variability, excel at encoding high-frequency stimuli through rate coding.

Objectives:

Simulate the behavior of regular and irregular vestibular afferents on neuromorphic hardware, building upon previous work [4].

Investigate the impact of head rotational inputs on these neuron populations.

The ultimate goal is to replicate the vestibular system model on the Dynap-SE1 neuromorphic chip. Regular afferents can be emulated using a small population of neurons, while irregular afferents necessitate an excitatory-inhibitory network to model intrinsic noise. This will create a stochastic network for irregular neurons to capture their higher variability and uncorrelated noise. The system will be evaluated using a random broadband head velocity stimulus simulating natural head movements.

A simplified test case involves one-dimensional head movement (left and right), modeled as a sinusoid where head speed is encoded in the sinusoid frequency. To input this into the chip, two approaches are considered:

Encoding the sinusoid as a constant input bias to the neuron (a novel approach).

Using rate coding to map the amplitude to an input firing rate.

 

Expected Outcomes:

A functional model of the vestibular system afferents on neuromorphic hardware.

New spike encoding

 

Requirements

Programming skills in Python, interest in neuronal systems (biological or artificial).

 

Reference

[1] Purves D, Augustine GJ, Fitzpatrick D, et al., editors. Neuroscience. 2nd edition.

[2] Corradi, F., Zambrano, D., Raglianti, M., Passetti, G., Laschi, C., Indiveri, G. (2014). Towards a neuromorphic vestibular system. IEEE transactions on biomedical circuits and systems, 8(5), 669–680. https://doi.org/10.1109/TBCAS.2014.2358493

[3] Sadeghi SG, Chacron MJ, Taylor MC, Cullen KE. Neural variability, detection thresholds, and information transmission in the vestibular system. J Neurosci. 2007 Jan 24;27(4):771-81. DOI: 10.1523/JNEUROSCI.4690 06.2007. PMID: 17251416; PMCID: PMC5053814.

[4] Chicca, E., Fusi, S. (2001). Stochastic synaptic plasticity in deterministic aVLSI networks of spiking neurons. In F. Rattay (Ed.), Proceedings of the World Congress on Neuroinformatics (pp. 468-477). Vienna: ARGESIM/ASIM Verlag.

 

Contact

Elisa Donati elisa@ini.uzh.ch