Current meter using GMR

 
I've come up with some ideas that use GMR
(http://en.wikipedia.org/wiki/Giant_magnetoresistance ) to measure
current flowing in a wire.  Does anyone have a similar idea ?

The idea is this, place a GMR device in the magnetic loop of a
transformer.  In one coil (low turns-large current) place the load and
the other - high turns low current place a feedback current.  The
feedback control circuit is designed to keep the GMR resistance at a
maximum.  Then you can measure the current (voltage across resistor)
flowing in the feedback coil and that should be directly proportional to
the current flowing in the load coil.

The issue is that the feedback controller will not know if it has to
increase of decrease current because of the symmetric nature of the GMR
device response curve.  For this you can use multiple GMR devices and a
special transformer geometry that will indicate whether the feedback
current needs to be increased or decreased.

Any thoughts ?