The best way would be a differential FET probe. I believe the the ATX standard pretty much specified a Tek P6046 for noise and ripple measurement. Something like a 100 MHz differential probe that goes down to 1 mV/div with a CMRR of 10k:1. Not sure what the modern replacement would be, do any of the low voltage 1x differential probe have AC coupling? I guess you could use a clean voltage source to provide an external offset.
The ATX standard for power supply testing specified a P6046 and a couple days ago I posted long post explanation as to why and what could be done with more modern equipment but the forum silently deleted it. I am not sure if that is a forum bug or if it was do to moderation.