X-Message-Number: 3482
From:  (Thomas Donaldson)
Subject: CRYONICS:re Keith Lynch's mapping scheme
Date: Tue, 13 Dec 1994 11:27:19 -0800 (PST)


There are lots of minor problems with the proposal by Keith Lynch for brain
mapping, but so far as I can see every one can be fixed. Lynch's proposal
also takes excessive effort in the mapping, with the result that an even
simpler system should still work (these problems come from a RELATIVE un-
familiarity with how brains work).

I will discuss some of them later, but the main point is that the basic idea
should work. It may also be much closer than the development of nonbiological
devices to do this work at nanoscales; as I mentioned, herpes virus is used
right now to trace out nerve connections. Herpes is useful because it will
propagate through existing nerve synapses. To use modified herpes virus,
we genetically modify it to do more than the existing ones --- this is a more
extensive modification than we now do, but abilities to do this kind of 
genetic engineering continue to grow.

In other words, a similar scheme to that of Lynch will arrive much sooner
than the devices he used to implement it.

A basic description of the highly modified Herpes virus would look like this:
it will create another set of copies of itself every time it crosses a 
synapse. These copies will contain a set of nucleotides coding their history
ie. who their parent virus was. They will also absorb information about the
synapse at which they were created. At the end of this process, the viral
particles become quiescent. Each will have a tag to be used later in data

Data collection would work as follows: other viruses will be injected. These
spread through the nervous system as before; whenever they reach a nerve cell,
they induce it to release all molecules bearing the tag above. The former
"mapping" viruses are thus extruded from the cell, picked up in the blood, 
and (say --- lots of possibilities exist ---) excreted in urine. The second 
set of viruses will be devised so that they are sensitive to a particular 
drug, nontoxic, and after doing their job they can be eliminated by this

1. Herpes does not require the neuron to be active in order to pass through
   a connection. What we really want here isn't the activity of a synapse,
   but its composition in terms of nerve transmitters and those modifications
   (presently unknown but likely to be soon found) which increase or decrease
   its propensity to fire.

2. Most nerves have several thousand connections. This will not significantly
   change the information requirement unless we insist (which we need not do)
   that the entire result be stored in the body of a subject. It's also 
   true that some neurons synapse on themselves and others can have more 
   than one synapse with the same neuron. Our coding scheme should be able
   to record these cases too.

3. There is no need for a clocking signal and hence no need for radio or
   microwave transmission/reception. 

4. Given present discoveries about our brain, we may want to map glial cells
   also. An extension of this scheme should handle them too. 

Finally, I believe that we would want to store not just the structure of our
brain but also our genome. This would allow us to recreate our body. Much
more important here, our brains don't operate autonomously, but in response
(among many other things) to our hormonal state. This is not determined by
our brain connectivity. And it will be complex, enough that rereading us
into any other form should have to take account of it. That is, our brain
alone does not provide complete information about our self --- but saving
our genome, after all, is far easier. I am just saying that we should
remember to do so.

A good reference on neurons for those who want to develop nerve mapping further
is Levitan and Kaczmark, THE NEURON.

				Long long life,
				   Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=3482