Unicompartmental knee arthroplasty MRI: impact of slice-encoding for metal artefact correction MRI on image quality, findings and therapy decision

To evaluate the impact of slice-encoding for metal artefact correction (SEMAC) on image quality, findings, and therapy decision in patients with unicompartmental knee arthroplasty (UKA). Forty-five painful UKAs were examined at 1.5T-MRI (STIR, proton-density(PD)-weighted sequence, each with SEMAC and high-bandwidth). Artefact size, image quality, anatomic depiction, and clinically relevant findings were compared between SEMAC and high-bandwidth (2 readers). In 30 patients, therapy decision was retrospectively assessed by two orthopaedic surgeons without MRI, with high-bandwidth-MRI, and with SEMAC-MRI. SEMAC reduced mean artefact size for STIR (11.8 cm² vs. 37.7 cm²) and PD (16.8 cm² vs. 18.9 cm²), p < 0.0005 for both comparisons. SEMAC showed more blurring than high-bandwidth, p < 0.0005. STIR-SEMAC revealed more bone marrow oedema (29 vs. 18 patients, p = 0.001, 30 vs. 13 patients, p < 0.0005, for reader 1 and 2 respectively). PD-SEMAC was worse in detecting meniscal lesions (6 missed, p = 0.031, 9 missed, p = 0.004, by reader 1 and 2 respectively) than PD-high-bandwidth. Revision-surgery was chosen in 12 and 11 patients without MRI (surgeon 1 and 2), with high-bandwidth-MRI in 15 and 14 patients, and with SEMAC-MRI in 19 and 14 patients. STIR-SEMAC was useful in detecting bone marrow oedema and influenced the orthopaedic surgeons' decisions towards surgery, while PD-SEMAC showed no clinical benefit. (orig.)