On Friday, June 6, 2003, at 08:26 AM, Thomas Shaddack wrote:

On Wed, 4 Jun 2003, Trei, Peter wrote:

...

It appears that they can't tell the medical isotopes from others....

They have no chance to distinguish isotope type with just a plain Geiger. For an identification, they would need a gamma spectrometer, which is a toy that AFAIK is not yet portable and cheap enough for mass deployment.

I certainly never implied in any way that a simple G-M tube would be useful for this. Implicit in my radioistope mapping comment was that a gamma ray spectrometer would be used. As for portability, the one I used in my lab in 1979-82 was not terribly heavy. The heaviest part was the LN dewar, which was large and floor-standing. A large dewar is certainly not needed. The rest of the assembly, even 20 years ago, was mostly portable: the germanium detector head, some preamps and pulse-height analyzers, and a multichannel analyzer. Most of this stuff is now done on laptops, the MCA and analysis software part. Without researching this on the Net, I would thus conjecture the entire gamma ray spectrometer could fit in a small carry-on case, using a small dewar. Certainly for the cost of operating a light plane, such a spectrometer would be a minor cost by comparison. And note that this is just what can be easily bought on the open market...N.E.S.T. (Nuclear Emergency Search Team) and similar LEO people almost certainly have more miniaturized detector setups. I expect most of the N.E.S.T. detectors are also gamma ray spectrometers, probably now so portable they fit unobtrusively into briefcases for use in crowded areas. As we discussed a few months ago (and I think I discussed this in _particular_ with Thomas!), the S/N advantages of using a spectrometer are enormous. Thousand-to-one improvements in general S/N are easily achievable. Even more if the MCA software is looking for pairs or triples or n-tuples of gamma peaks and inferring likely radioisotopes. (I used this approach in 1981 to solve a major problem in IBM computers which were using Intel chips...and I don't mean the alpha particle soft error problem. This was a different problem, involving a beta source trapped in some of the packages. For this I used a pair of large sodium iodide crystals (which my well-equipped lab just happened to have in a storage cabinet, fortunately for us) and looked for a specific decay mode that resulted in a pair of gammas sent out in opposite directions. By using coincidence logic over microsecond intervals, enormous improvements in S/N could be achieved. Basically, background radiation vanished and only the specific beta decay mode we were looking for appeared.) --Tim May --Tim May "Extremism in the pursuit of liberty is no vice."--Barry Goldwater