There is a page on the 40m wiki explaining the procedure.
Additionally there are several old elogs about the cross-coupling minimization, which can be useful for us:
 "SUS ranking from measured data", iLog by Osamy (Aug.22 2005)
 "TF from position to sensors for ETMX ", iLog by Osamu (Aug.25 2005)
 "Further OSEM tweaking", iLog by Rana (Oct.3 2006)
We need a plan how to minimize cross coupling in the OSEMs now