poor man's uploading

Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Fri, 20 Sep 1996 12:58:10 +0200 (MET DST)


I volunteered for a MRI brain scan yesterday... Both anatomic and
functional. I'm going to stream the voxel data set onto a DAT cartridge
to toy around with at leisure. Medical image processing, especially on
voxels is currently making giant strides forward.

Just a few impressions: you don't notice a (comparatively) low-Tesla
field, unless you have something ferromagnetic. I've been told in 3..4 T
research imagers you feel the induced electricity in terms of jolts/spasms.
I've waved my hand vigorously in front of the magnet & felt absolutely
nothing...

Concerning ferromagnetic objects: in older MRI's you felt the pull at
your keys/wallet already at entering the tomograph room. It has become
better, due to active shielding. If a coin escapes into the tube,
you can usually slide it outside & then detach it. Pictures become bad if
a coin is in the tube, though. Field's becoming inhomogenous.

Larger pieces of steel are downright dangerous. I've been told of a
video: there a modest-sized pipe wrench experienced a 70 kg (sic) pull.
For this particular object this meant a 10 g acceleration. You are
lucky, if the thingy stays in the tube, usually it flies right through,
gets deccelerated as it leaves the tube, returns & comes back sailing
out. At you. Not recommended.

In one documented case a floor-washing machine has been yanked into the
magnet... I shudder to think what would happen if you happened to stand
between the magnet & the machine at that moment... Great fun.
Squashed human sandwich.

You hear the different pulse sequences: the gradient coils experience a
force in the strong field. Rectangular pulses sound harsh, sinus is
downright melodic. With some training, you could probably write down the
pulse sequence from hearing ;) . The entire scan lasted for about 2.5 h.
It somewhat resembles an isolation tank experiment: you can't move, you
see only the inside of the tube, hearing only extremely monotonous sounds.
(It's pretty loud, one has to use oropax). Apart from experiencing acustic
hallucinations, you grow really sleepy really soon.

Though UKW power is not negligeable, you don't feel warmth in a head
scan. With body scan this is supposed to be different. On larger-Tesla
machines some pulse sequences are locked, since they would induce
harmful potential deltas. It's not funny if you're downright obese as
well. Then you might feel it even in standard MRI.

Functional MRI uses the fact that oxy/desoxy hemoglobine has different
magnetic properties. In a motorics experiment, moving your fingers
requires activity in the according motoric homunculus area of the cortex,
which needs stronger blood flow, etc. Temporal resolution's bad, but
spatial is very good. In contrast to PET, this is both cheaper & less
hazardous (both for the environment & for you).

During the scan the entire room is full of heavy EM smog. It would
probably roast the average microcontroller, so they use fiber optix.

For physical reasons, minimal voxel size is several microns, and then
only for tiny specimens (insects/mice). This is called MRI microscopy.
For uploading, vitrified cerebra are best. Unless incremental in
vivo uploading is impossible, it has to be destuctive scan time...

ciao,
'gene

P.S. Sorry for the private mails I didn't yet answer: they're in the
processing pipeline.

_________________________________________________________________________________
| mailto: ui22204@sunmail.lrz-muenchen.de | transhumanism >H, cryonics, |
| mailto: Eugene.Leitl@uni-muenchen.de | nanotechnology, etc. etc. |
| mailto: c438@org.chemie.uni-muenchen.de | "deus ex machina, v.0.0.alpha" |
| icbmto: N 48 10'07'' E 011 33'53'' | http://www.lrz-muenchen.de/~ui22204 |