Clinical MR imaging of patients with spinal hardware at 0.55T: comparison of diagnostic assessment and metal artifact appearance with 1.5T.

PURPOSE

The aim of this study is to assess inter-reader agreement of imaging findings and compare readers' assessment of image quality (IQ) and appearance of metal artifact (MA) in patients with spinal implants between 0.55T and 1.5T MRI.

METHODS

Patients imaged on Siemens Healthineers Magnetom Free.Max 0.55T (n = 42; avg. age 55 yrs.) with spinal hardware between 12/2021 and 3/2024 were included. Of these, 18 patients had a paired exam at 1.5T. All exams were reviewed independently by three neuroradiologists (R1-3). Readers selected imaging findings from a pick-list and rated sequences using a 4-point Likert scale for IQ and MA.

RESULTS

At both 0.55T and 1.5T, raw agreement for the following findings ranged between 81 and 95%: cord signal abnormality, osteomyelitis/discitis, osseous metastatic disease, and compression fracture. Agreement on post-operative fluid collection and spinal canal stenosis was 64.3% and 66.7% at 0.55T, and 77.7% and 50.0% at 1.5T. Agreement on neural foraminal stenosis was low in both cohorts, 47.6% and 33.3% at 0.55T and 1.5T. No sequence at 0.55T was rated inferior to 1.5T in IQ or MA. Sequences rated higher at 0.55T compared to 1.5T were as follows for IQ: sagittal T1w TSE (R1) and as follows for MA: axial T1w TSE (R1, R2), sagittal T1w TSE (R1), axial T2w TSE (R1), sagittal T1w TSE Dixon post-contrast (R2), sagittal T2w STIR (R2: p = 0.01).

CONCLUSION

Imaging patients with spinal hardware at 0.55T results in comparable inter-reader agreement for clinically-relevant imaging findings and equivalent or improved image quality compared to 1.5T.