Lorenzo Nardo MD, Presenter: Nothing to Disclose

Misung Han, Abstract Co-Author: Nothing to Disclose

Sonia Lee MD, Abstract Co-Author: Nothing to Disclose

Ursula Renate Heilmeier MD, Abstract Co-Author: Nothing to Disclose

Kevin Koch PhD, Abstract Co-Author: Employee, General Electric Company

Thomas M. Link MD, PhD, Abstract Co-Author: Research Grant, General Electric Company Research Grant, InSightec Ltd

Roland Krug PhD, Abstract Co-Author: Nothing to Disclose

Given its higher signal-to-noise-ratio high field MRI at 3 Tesla is increasingly used for musculoskeletal applications, however, metal artifacts and related image distortions are also more pronounced at 3.0T. We therefore studied the impact and the feasibility of metal artifact reduction sequences at 3.0T as compared with 1.5T

Twenty patients (aged 58-76) with total hip replacements were scanned at 1.5T and 3.0T within a period of 2 weeks. For signal acquisition, 8-channel phased-array cardiac coils were used on both scanners. The sequence protocol included: multiacquisition variable-resonance image combination (MAVRIC) PD (coronal), MAVRIC STIR (axial) as well as standard FSE sequences. Each study was assessed by 2 radiologists for morphological abnormalities (joint effusion (including findings suggesting aseptic vascular autoimmune lymphocytosis (AVAL)), bone marrow edema pattern, osteolysis, insertion tendinopaty at the greater trochanter) and distinction of anatomic details (anterior and posterior femoral head, femoral neck, greater and lesser trochanters). A five-point scoring system was used: 1- good visualization, 2- good visualization with minimum artifacts, 3- visualization not compromised by artifacts, 4- visualization compromised by artifacts, 5- severe artifacts. Furthermore, the extent of artifacts was quantitatively measured. Wilcoxon signed rank test was used to compare the data obtained by the two different scanners. Agreement between the two readers was reported with kappa values.

While the extent of artifacts was significantly smaller at 1.5 T compared to 3.0 T (p0.05): average scores ranged between 2.5 and 3.4 at 1.5 T and between 2.6 and 3.3 at 3.0 T. Also the assessment of morphological abnormalities was not significantly different between the two field strengths (p>0.05) with average scores ranging between 2.6 and 3.5 at 1.5 T and 2.5 and 3.6 at 3.0 T. Inter-reader agreement for different anatomic details and clinical findings visualization ranged between k=0.65 and k=0.90.

Though artifacts were larger at 3 T compared to 1.5 T, the visualization of morphological abnormalities and anatomic details was not significantly different between the two field strengths.

3.0 T metal artifact reduction showed no significant difference in detection of morphological abnormalities and anatomic details when compared to 1.5 T.

Nardo, L, Han, M, Lee, S, Heilmeier, U, Koch, K, Link, T, Krug, R, Is Accurate Metal Artifact Reduction Feasible at 3T? A Comparison Study between 3T and 1.5T.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13044424.html