Purpose: This study was performed to assess the feasibility, interobserver variability, sensitivity, specificity and diagnostic accuracy of raw data and postprocessed images from a low-field (0.5-T) magnetic resonance (MR) unit in evaluating vascular complications of kidney grafts.
Materials and methods: We enrolled 49 patients undergoing MR angiography (MRA) for a clinical suspicion of renal artery stenosis. The raw data, maximum intensity projections (MIP) and multiplanar reconstruction (MPR) image sets were evaluated independently. We calculated the number and degree of stenosis, and sensitivity, specificity and accuracy for MIP, MPR and raw data image sets, together with interobserver variability.
Results: Interobserver agreement was substantial. There were no differences among the MIP and MPR algorithms and raw data images for the detection of stenosis. Raw data images were more accurate in quantifying the severity of stenosis, with higher sensitivity (75% vs. 62.5%), equal specificity and higher diagnostic accuracy (75% vs. 71.43%).
Conclusions: Contrast-enhanced MRA, even with a low-field (0.5-T) unit, is a feasible, sensitive and accurate technique for the study of the renal arteries of the transplanted kidney.