ARRS 2019: Radiofrequency heating of hip implants during 3T MRI

Clinical magnetic resonance imaging (MRI) protocols used for metal artifact reduction sequences (MARS) when imaging patients with hip arthroplasty implants pose only minimal risk of causing thermal injury, according to a study presented at a scientific session of the American Roentgen Ray Society’s (ARRS) annual meeting held May 5-10 in Honolulu, HI. Researchers reported that the maximum heating of a variety of different hip implant constructs by a variety of clinical MARS MRI pulse sequences at 3 Tesla (3T) was less than 2°C.

Lead author Iman Khodarahmi, MD, a radiologist at NYU Langone Health in New York City, explained that two rectangular American Society for Testing and Materials (ASTM) MRI phantoms were placed head to head on a MRI scanner table to simulate the upper and lower portions of a human torso. They were filled with gelled saline medium which had the electrical and thermal properties of human muscle.

The researchers tested the phantoms with four different total hip arthroplasty implant configurations. These included two metal-on-ceramic constructs with titanium femoral stems at two lengths, a metal-on-metal construct with cobalt chromium femoral stem, and a metal-on-polytethylene construct with cobalt chromium femoral stem. Each implant was mounted inside the phantom on non-conducting plastic holders at the location of maximum heating inside the bore of the scanner.

The researchers measured temperature changes of three clinical pulse sequence type protocols using fiber optic temperature sensors with ± 0.1° C accuracy. These included high-bandwidth turbo spin echo (HBW-TSE), slice encoding for metal artifact correction (SEMAC), and compressed sensing SEMAC (CS-SEMAC). Each protocol contained six pulse sequences obtained in coronal, sagittal, and axial orientations as intermediate-weighted and short tau inversion recovery varieties with image coverage identical to that in patients.

A 30-minute HBW sequence was modified twice to produce high, “supraclinical,” specific absorption rate (SAR) values to assess potential extreme scenarios. These were not clinically permissible protocols, because they exceed the hard-coded U.S. Food and Drug Administration (FDA) limits on whole or localized body SAR and/or scanner hardware protective thresholds.

“The maximum heating at any single point was 1.9°C in clinical sequences,” Dr. Khodarahmi reported. “Maximum temperature rises occurred at the tip of the femoral stem and medial aspect of the acetabular cup in most cases. There was no significant heating difference among various points along the implant periphery, and the degree of heating was not different among different implant types.”  He added that heating increased to 3.6°C with the supraclinical high SAR pulse sequences at up to 30 minutes of continuous MRI.

“The results suggest only a minimal risk of thermal injury with the clinical MARS MRI protocols used, especially when adding the cooling effect of perfusion in physiologic conditions,” he concluded.