Minimizing RF heating of conducting wires in MRI

Abstract

Performing interventions using long conducting wires in MRI introduces the risk of focal RF heating at the wire tip. Comprehensive EM simulations are combined with carefully measured experimental data to show that method-of-moments EM field modeling coupled with heat transfer modeling can adequately predict RF heating with wires partially inserted into the patient-mimicking phantom. The effects of total wire length, inserted length, wire position in the phantom, phantom position in the scanner, and phantom size are examined. Increasing phantom size can shift a wire's length of maximum tip heating from about a half wave toward a quarter wave. In any event, with wires parallel to the scanner bore, wire tip heating is minimized by keeping the patient and wires as close as possible to the central axis of the scanner bore. At 1.5T, heating is minimized if bare wires are shorter than 0.6 m or between ≈2.4 m and ≈3.0 m. Heating is further minimized if wire insertion into phantoms equivalent to most aqueous soft tissues is less than 13 cm or greater than 40 cm (longer for fatty tissues, bone, and lung). The methods demonstrated can be used to estimate the absolute amount of heating in order to set RF power safety thresholds. © 2007 Wiley-Liss, Inc.