[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Falso Positives on TLD



After a coupla inquiries this morning, I've condensed and organized my longer
message (which I didn't post to RADSAFE because it was so long) into the
shorter (and hopefully more readable) message which follows. Hope it helps.

Tony A. Rhea, CHP, HP Manager
Health Protection Division
SAIC / Oak Ridge, TN
tony_rhea@cpqm.saic.com or 615/481-2345
***************************************************
Stu Torf (torfsm@nu.com) of the RP section @ Northeast Utilities writes:

>>Has anyone experienced "false positive" beta responses with 
>>their TLDs? We  have performed glow curve analyses on these 
>>responses which say that there is a real beta response but 
>>our TLD users say that there is no beta. Any comments?  

Yes, we have looked at this problem extensively. A full discussion of it may
be lengthy, however, and may be better suited to a phone call. However, some
food for thought follows to get the conversation started.
 
Relevant factors are your LLD/"administrative reporting limit," your
algorithm, your badge design (I think NU uses the Harshaw 8800 series -- do
you have the 3.6 or 6 mil E3?), your reader stability, etc. How do your RCF
control charts look? What is your deep dose LLD from gamma? How variable is
your background data (that is, how big are the error bars on the data points
of a plot of background badge response -- per element -- versus time). How do
you correct for / calculate background? Do you use a linear assumption (y = mx
+ b, or response = ndays * rateperday + constant -- for each of the four
elements). What kind of window do you have over the "thin chip?" Is it black
mylar or plain mylar. Have you done a UV study?

Some places use a "post-algorithm filter" to screen out calculated results <
their "administrative reporting limit." Note the distinction between LLD and
admin reporting limit -- the former is a formula whereas the latter is an
administrative / policy decision. What limits, policies, and practices do you
use?

Hope this helps get you started. Call if you want to talk in detail.
***************************************************