How much bandwidth does the spinal cord have?
11/08/2018
From Paulingtons on Reddit:
This is an interesting question, if not near impossible to answer properly. However I figured I'd give it a go even if I do have to make some gross assumptions.
First, we need to know how many neurones are in the spinal cord. That's very hard to know, unless we make some assumptions.
The spinal cord diameter is variable, from the small ~7mm in the thoracic area to the ~13mm in the cervical and lumbar intumescentia (enlargements), let's average that out to 10.5mm in diameter. It is also not a perfect circle, but let's ignore that for now.
Now the diameter of an axon is similarly difficult, they range from one micrometer up to around 50 micrometres, with far more in the <5 micrometre range. However a study found that the average diameter of cortical neurons was around 1 micrometre D. Liewald et al 2014 plus 0.09 micrometres for the myelin sheath, so let's say the average diameter of a neuron is 1.09 micrometres.
Okay, so let's simplistically take the area of the spinal cord (Pi * 0.01052) and the same with the neuronal diameter and we get:
( 7.06x10-4 m2 / 3.73x10-12 m2) = ~200,000,000 neurons in the spinal cord.
Now, given that there are around ~86 billion neurons and glia in the body as a whole, with around ~16 billion of those in the cortex (leaving 60 billion behind) I would wager that my number is an underestimate, but let's roll with it.
Okay, so we know how many we have, so how fast can they fire? Neurones have two types of refractory periods, that is absolute and relative. During the absolute refractory period the arrival of a second action potential to their dendrites will do absolutely nothing, it cannot fire again. During the relative refractory period, a strong enough action potential could make it fire, but it's hard.
So let's take the absolute refractory period for an upper limit, which is around 1-2ms Physiology Web at the average of 1.5ms. This varies with neuron type but let's just roll with it.
So we have ~200,000,000 neurones firing at maximum rate of 1 fire per 0.0015 seconds. That is ~133,000,000,000 signals per second.
Let's assume that we can model neuronal firing as "on" or "off", just like binary. That means this model spinal cord can transmit 133 billion bits per second, and a gigabit = 1 billion bits, which gives our spinal cord a maximum data throughput of 133 gigabits per second.
Divide that by 8 to get it in GB, and that's 16.625 GB of data per second capable of being transferred along the spinal cord. Or about a 4K movie every two seconds.
DISCLAIMER: This is all obviously full of assumption and guessing, think of it as Fermi estimation but for the spinal cord. It's not meant to be accurate or even close to being accurate, just a general guess and a thought experiment, more than anything.
Source: Neuroscience student.