## Control and Feedback in the Central Nervous System

From: andrew cooke <andrew@...>

Date: Fri, 5 Jul 2013 20:02:51 -0400

[Before I start, a disclaimer - this is just my attempt to place a structure
on recent experiences.  I am not a medical expert, at all.]

A single "cycle" of (recurrent) MS consists of a short "outbreak", when the
immune system attacks your nervous system, followed by a much longer recovery
phase, during which many of the nerves are repaired.  Experiencing this is,
amongst other things, interesting, because it gives you some insight into how
the central nervous system (CNS) works.

So what I will do below is list various experiences, and then try to link
things together, into a coherent description.

1 - You don't feel much when things go wrong.  The best analogy I have for the
damage done to nerves is that it's similar to removing the insulation from
electrical cable.  In an electrical system that likely gives you noise -
in a biological system what you feel is pins and needles (without the
painful part that seems to be part of pins and needles when you restrict
blood flow to your leg, for example - it's just the tingly feeling).

In addition, you feel various muscles as "lumps" or "sausages" inside your
body.  As far as I can make out this is the best interpretation that the
conscious mind can make of inconsistent signals, when some (working)
nerves detect the presence of a nearby muscle, but the (damaged) nerves in
that muscle are silent.

But what you don't feel is a gap.  There's no feeling of "whoa, I can't
feel X", unless you are making some conscious effort to sense something.
So if your fingers are numb, and you rub your hand across a rough cloth,
you might think "huh, I can't feel that".  But otherwise, there's no "my
fingers aren't sending the usual data" feeling.

2 - Sensing and control seem to be separate.  Damage to the CNS occurs in
specific locations (lesions).  So if that's physically where nerves for
your arm are, then your arm is going to lose both control and feeling.
But at a finer scale, feeling and control seem to be decoupled.  Some
parts of your arm might be numb, but still move just fine, for example.

3 - There's redundancy and overlap.  One of the more unusual experiences as
nerves in your legs recover is hyper-awareness that you're wearing
trousers.  When nerves that feel their presence "turn on" again, the
signal is initially way too loud (presumably because it is new; it decays
to normal levels in a day or so).  So you are over-aware of these hot,
restrictive, uncomfortable nappy (diaper)-like things.

And this (unmistakable) feeling occurs multiple times.  Which, as far as I
can see, means that multiple circuits are basically detecting the same
thing.

4 - For a long time (~3 months) I was unable to run / jog for more than 15
minutes.  After that, I was completely exhausted.  Not physically
exhausted - more a weird kind of emotional / mental exhaustion.

Then, one day, it went away, and I could run further.

5 - The process of recovery takes a long time (for me, things are still
improving after 8 months), but appears mainly as very localized
improvements: single muscles aching or twitching; localized pins and
needles; return of feeling in a small area; appearance or disappearance of
a "sausage"; etc.  The "big" improvements, like being able to walk up and
down stairs smoothly, or being able to run for more than 15 minutes, are
much more infrequent (and seem to be sudden, rather than gradual).

6 - You want to sleep at various points during the day.  This seems to be a
very common symptom.  I was already good at napping, so it's not really
bothered me, but at least twice a day I will spend 15-20m asleep (before,
you feel really tired; after, you feel great).

OK, so here comes the synthesis...

All the above seems to fit well with a two layer model.  The "upper" layer
would be a negative-feedback system - if you want to move your legs, it sends
out commands, then receives back reports, calculates the error, and corrects.
The "lower" layer is similar but learns (or learnt, back when you were first
trying to walk) the appropriate commands and offers "open" (no feedback)
replays of standard motions.

From the amount of "small changes" experienced, and long recovery period, it
seems like there was a lot of damage to the nerves.  Yet my "large scale"
symptoms were not that bad - I could still walk OK, use both hands (although
somewhat crudely - writing was a mess, tying laces tricky).

My guess is that the redundancy and overlap, plus the upper layer of feedback
control, are enough to keep your body working pretty well, even when there's
pretty major damage.

But there's a cost.  Even though that is a subconscious process, it seems to
require significant effort.  Hence the tiredness (both daily, and when
running, which presumably requires extra coordination).

Recovery is through many small, uncoordinated fixes.  At some point they are
sufficient to allow the lower level - replaying learnt sequences with little
feedback - to kick in.  And that's when I could suddenly run further, without
exhaustion.

In other words:

* Redundancy and feedback keeps things working, at a cost.

* Recovery is progressive, but makes little difference to "macro" symptoms
until a threshold is crossed, at which point learnt actions work again.

Andrew