X-Message-Number: 28547 From: Date: Tue, 3 Oct 2006 16:21:05 EDT Subject: Uploading technology (1.iii.3) MRI Brain Reader 3 Uploading technology (1.iii.3) MRI Brain Reader 3 In messages .#27635 and #27639 I had given a first look at an MRI brain reader. Here I would show a specificity of such a machine. It can't be on the track of current medical systems. There is no hope it will be built as the technology evolves, it is another branch of the tree. Assume that the depolarization time T1 is one millisecond in ice, this is an order of magnitude smaller than for the liquid phase. The recovered signal will be stronger than with liquid state T1 on the order of the second or more. This is a good thing because all the energy is releaser in a short time giving off more power and so a better signal over noise ratio. When there is a microscopic imaging, a difficulty arises. Assume the signal has the maximum duration, that is, it is equal to T1 and is one millisecond long. We can think of it as a monochromatic frequency modulated with all frequencies between zero and one kilo Hertz ( 1 kHz has a cycle duration of one millisecond). If He3 is used, its Larmor precession frequency is 3.26 kHz/gauss. So, to discriminate from one pixel to the next, nearly .3 gauss are called for, so that the frequencies are more than one kHz appart. If the resolving power is 100 nanometers, the magnetic field gradient is .3g/100nm or, for a 4 inches picture line : 300 000 gauss or 30 Teslas. The continuous field must be at least 3 times larger, so the homogeneous field is 100 T. This is at least 3 times the best laboratory limit. Most of the grey matter in the brain containing neurons are in a shallow coating no more than 2 to 3 millimeters thick. Even without surgery, this could be reached with one centimeter gradient. We are down to 10 T for the continuous field and 3 T for the gradient. This is the technological limit for small volumes. We could picture a rodent brain, no more. Rodents have a small brain and few convolutions without of the reach parts. Man is another story. The NMR is so not up to the job. But the title is about Magnetics Resonance Imaging, not Nuclear Resonance Imaging. MRI includes both, nuclear and electronics magnetics resonance. Electrons are 2000 times lighter than hadron (protons, neutrons) in the nucleus. So, they are swinging on the order of 2000 times faster. The Larmor frequency is now in the MHz range per gauss. Assume it is 10 MHz/gauss for a given electron. A 100 nm resolution needs always 1 kHz/pixel, but that translates into .0001 gauss/pixel or one gauss/mm. If the resolving power is pushed to 10 nm and the picture is one feet wide, the gradient field is only 3,000 gauss. The continuous field must be at least x3 larger or near one Tesla. This is perfectly in the range of present day technology. For such a field, there is no need for exotic technologies as superconductor in liquid helium or the like. Electron Resonance Imaging is the solution. YB. Content-Type: text/html; charset="US-ASCII" [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=28547