Sirous,

I think the confusion may stem from the IRN you are looking at.  In the circuit, the inputs are currents, so an IRN measured in voltage will not be appropriate.  You should measure the IRN in something like pA/√Hz.  That may help.  Also, it looks like your circuit is maybe some kind of sense amp for a RAM?  If so, IRN is probably not the most convenient way to specify noise performance.  You may want to think of the noise as almost a random offset current, since that will help you to visualize it.  generally, in a sense amp, you have prior knowledge of how much the input signal will swing (current in this case) so it isn't hard to specify a 3-sigma noise value.

Good luck,
Carl