X-Message-Number: 3978
From: "Joseph J. Strout" <>
Newsgroups: sci.cryonics
Subject: Re: LifeScan2
Date: Wed, 8 Mar 1995 08:58:19 -0800 (PST)
Message-ID: <>
References: <>

On Tue, 7 Mar 1995, Eugen Leitl wrote some thoughtful and useful comments 
on the functional scanning of vitrified brain tissue:

> 0) to read the neuron type (susceptibility function)
> 1) to trace the tubular structures of neuron's axons
>    and cell body (graph connectivity, edge weight)
> 2) distinguish them from the background (glia cells)
> 3) read the synapse value and sign (weighted edge 2)

> To 0) Whether this is discernible from shape alone or
> the membrane proteins type.. No one knows.. (BTW, there
> are some 50 types of neurons.)

It is is not clear to what extent the many "types" of neurons are 
functionally distinct types of neurons apart from their morphology.  That 
is, most of the divisions are made on the basis of "well, this one looks 
like a starfish, and that one looks like a pyramid..."  The morphology 
has obvious effects on synaptic integration etc. (more below), and is 
easily captured by a detailed morphological scan.  The other properties 
which vary from type to type (intrinsic firing properties, etc.) may be 
deducible from the combination of morphology and location within the 
brain -- a given region often has only a few types of neurons.  On the 
pessimistic hand, it may be necessary to determine subtle differences in 
gene activation and protein content.  Not an insurmountable problem, but 
a difficult one.
 
> To 1),2): One can view the tubular lipid membrane bilayer
>
>		(...reasonable description of axons omitted...)

> In short, axons alone can be approximated/modeled by weighted edges
> (graph theory lingo), the weight being the delay. Propagation
> velocity is a global constant. (Thanks god.)

The former statement (axons as weighted edges) may oversimplify 
overmuch.  For example, axons which run in parallel can influence each 
other by their sharing of extracellular ions.  A better example is the 
many varicosities where axons make "en passant" synapses, or the branch 
points which act as low-pass temporal filters (rapid impulse trains are 
dampened, while slow ones are unaffected).

The second statment -- propagation velocity is a global constant -- is 
incorrect.  Propagation velocity depends on the ration of internal to 
external resistance, i.e., the axon diameter and degree of myelination, 
which varies widely.

> Applying 3d image processing and recognition techniques (edge detection
> and tracing) upon voxel blocks (critical maximum size determined by
> the processing systems' memory size) makes for a high hope for 
> easy/successful axon scan/tracing/modeling. 

This sort of work is (slowly) getting underway.  The term "large-scale 
ultramicroscopy" (LSUM) describes the process of imaging & encoding data 
at an ultrastructural level over wide areas (a cubic millimeter is a 
reasonable goal for the next decade or so).  Right now, the major 
bottleneck seems to be in image processing -- automatically recognizing 
relevant structures in each slice, and combining them appropriately to 
generate the 3-D representation.  Currently this is usually done 
painstakingly by hand.

> The state of the neuron is the probability of firing in the next
> instant or the overall (time slice) firing frequency. By applying several 
> transmitters (excitatory/inhibitory), which get depleted into
> the subsynaptic cleft by the synapse upon signal spike transit 
> and diffuse (through passive physical transport) to the postsynaptic 
> membrane of the neuron body and modify the probability of the 
> individual neuron to fire the next instant the whole
> of NN signal processing is done.

Hmm, this seems a good high-level description, though again it is 
probabla bit oversimplified for describing real neurons.  I think, at a 
minimum, you need to keep track of membrane potential at many points in 
the neuron, since it is this potential which is affected by synapses, 
integrates them, and (possibly) results in triggering an action 
potential.  It will probably also be necessary to keep track of ion 
ratios, channel densities, and gene expression, etc.  On the other hand, 
much of this cellular machinery may be "collapsable" logically to simpler 
terms... much research needs to be done to determine the minimal 
functionally accurate neuron model.

> What we might well expect that the dynamic range of a synapse
> is quite limited. While certainly not binary it should not 
> exceed 6 bits, tops 8 bit. The main difficulty in scanning is
> to attribute such a value to a tiny physical structure (say, 500-100
> nm sized) with a tolerable accuracy. Really tough.

Right.  It's not even clear how we would measure the synapse strength.  
Possibilities include (1) the number of presynaptic vescicles, (2) the 
area of the postsynaptic density, (3) the number and size of nearby 
mitochondria (a bit of a stretcher, but I'm brainstorming here...).  This 
is a question that might be well suited to low-density culture studies.  
Get a handful of cells (or better yet, just two) to grow processes and 
make functional synapses in a dish.  Impale the little guys with 
electrodes to measure the functional strength of the synapse(s).  Then 
slice 'em up, put 'em under the electron microscope, and see what info 
you can get to correlate with the measured synapse strength.  That would 
be a darn good study, which I don't think has been done yet.  Comments?

> P.S. These Random Ramblings shall be augmented by a 
>      real-world NN-model/encoding, together with
>      a suitable silicon/ME implementation discussion in the
>      next future.  

Good for you!  This is exactly what we need.  There is a lot of work to 
be done, and much of it can be done with modest tools and equipment.  
I've found several people now who have expressed an interest in actually 
doing research towards uploading on a small scale (let's call it 
micro-uploading, and imagine uploading tiny networks of cells such as a 
leech ganglion or a cultured network).  It may be time to start a mailing 
list for discussion of this sort of work.  All opinions welcome.

Joe Strout             Neuroscience, UCSD            

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=3978