Interpreting ELISA test results - (Jun/03/2010 )

Hi

I am wondering what is the relationship (mathematical, statistical, or otherwise), if such a relationship exists between the output luminescence/absorbtion and the antigen/antibody consentration in the test sample.
(Please excuse any ingnorance on my part, as this is not my field of training)

Thanks

Edit: Also is there a direct relationship between number of antigens/antibodies and the actual viral load (higher numeber of antigens represents a higher viral load) if an ELISA test used for HIV detection.

here is an example of plot for a cancer marker:

http://www.biocheckinc.com/inserts/bc-1017_ca19-9.pdf

here is example of curve for HIV ag p24

http://www.biotraces.com/Bio54.html

Thanks for your help