Speaker
Description
Layer-5 neurons send two signals down the same axon. Single spikes carry bottom-up input and bursts carry top-down input. I checked whether demyelination removes only the top-down signal. I simulated 1000 two-compartment neurons and sent their spike trains through a model of a myelinated L5 axon. I tuned the axon to published data, and it matches measured conduction velocities (1.09 m/s healthy, 0.37 m/s demyelinated) and burst spike failures (9% and 31%). I tested five decoders on how well they recover each signal (r²). For fast, in-vivo-like bursts, the bottom-up signal survived all lesions. The top-down signal was not lost but changed form: a linear decoder still recovered 93-99% of it, but facilitating synapses, which count spikes inside a burst, lost 32-42%. Myelin loss alone did little; it mattered only with damage at the nodes. Every lesion roughly doubled spike delay, so delay is the first sign.
Other Fields of Research (Track) for Abstract
Membrane Biophysics
| Nature of work | Computational |
|---|