X-Message-Number: 27639
Date: Tue, 21 Feb 2006 02:14:42 EST
Subject: Uploading technology (1.ii.3) Brain reader steps 2.

Uploading technology (1.ii.3) Brain reader steps  2.

The main magnetic field in an MRI scanner is  continuous. It is very 

difficult to get a high field producing a strong  polarization and at the same 
time a 
good homogeneity. I think it would be  interesting to produce a pulsed main 

field, may be up to 5 Teslas or even more  without too much smoothness and then
shut it down and let a weaker field with  very good homogeneity.

The first brain reader would be only a  "cracks reader",  a system looking at 
microcracks in the body. It could be  derived directly of the gas 

demonstrator and use liquid neon 21, the common  isotope of neon. It would work 
at minus 
240 C, a step colder than liquid  nitrogen. The cold temperature would bring a 
factor of ten in magnetisation and  so in received power. One atom out of 20 
000 would be polarized at 5  Teslas.

The antennas would be put in the neon bath and could be  build from high 
purity aluminum or berylium. The Q coefficient (a mesure of the  resonator 
quality) may be in the million range with them at these temperatures.  One 

possibility is to boost Q with an operational amplifier using the high  mobility

electron transistors technology. That system is used for satellite  reception 

One problem is the 1/f noise or "brown  noise", it dominate the electronics 
and thermal noise under 150 MHz.  Unfortunately, MRI above that frequency 

implies a very strong field. It could be  interesting to look at harmonics of 
main signal, one drawback, beyond the  reduced signal power, is the aliasing 
noise produced at high  frequency.

The strong coupling of atoms in a solid reduce both, the  T1 (spin-network 
relaxation time) and T2 (spin-spin exchange). T1 is above one  second in most 
liquids and gas, it is under the millisecond here. This is seen  as a strong 
drawback in the traditional view. Here, it is a good thing. A short  T1 will 

radiate all the energy in a short time, giving more power. Power, not  energy, 
what count in the signal/noise ratio. Getting a better S/N is very  important 
for microscopic MRI where voxels (volume elements) are  tiny.

The most interesting signals would come from phosphorous and  sodium, both 
are tracer of the membranes and neural receptors. Water (hydrogen)  gives a 
strong signal, but most molecules are not interesting. Carbon 13 may be  more 

useful,. the most common form of carbon, carbon 12, has no nuclear spin.  Both,
C13 and P31 are "classical" spin 1/2 systems, Na23 has a spin 3/2  structure.
 If chemical analysis can be included, by detection of the molecular  shift, 
then hydrogen would be more interesting, may be deuterium could be the  

target, reducing the interaction between nearby atoms. The D spin is 1 and the
Larmor frequency is : 654.5 Hz/Gauss.

It is common practice in two  dimensional MRI to code along X in frequency 

and along Y in phase. The process  is then inversed and a picture is taken with
frequency coding along Y and phase  along X. Without phase coding, a set of 
frequency coding in a number of tilted  directions can build back a picture. 

Using simultaneously the two process, that  is, taking a large number of  tilted
frequency-phase 2-d pictures could  recover the chemical shift "dimension". 
The third space dimension would be coded  in echo time by a time shift in the 
pi/2 radio pulse. Spin recovery sequence is  not a possibility here if the 

magnetization field is pulsed. This situation is  similar to the hyperpolarized 

Not all these elements  would be implemented at start, so this summarize a 
number of steps, may be ten  years of R&D.

Yvan  Bozzonetti.

 Content-Type: text/html; charset="ISO-8859-1"


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=27639