This work has introduced and evaluated a new method for metal artifact reduction and demonstrated its utility in routine clinical datasets. The greatest improvements were seen in dental fillings, where iMARv2 significantly improved image quality compared with conventional iMAR.