Reconstructed three-dimensional MR images: application to simulated tibial plateau depression fractures

Abstract

Purpose The aim of this study was to test the accuracy of measurements obtained from three-dimensional (3-D) reconstructed magnetic resonance images in the setting of a depressed tibial plateau fracture. Methods A simulated bilateral tibial plateau depression fracture was surgically induced in a disarticulated ovine knee. An attached perspex spacer was used to allow controlled fragment displacement, while maintaining a constant fracture gap. Ten identical intermediate-T2-weighted acquisitions (2.0 mm slice thickness) were acquired at different fragment displacement levels. Using a commercially available software package, these data were then used to generate 10 corresponding sets of 3-D images. Computer measurements were made of the volume of bony fragments, the width of the fracture gap, the degree of infero-medial and infero-lateral fragment depression and these findings were compared with caliper and volume displacement measurements taken from the phantom itself. Results Mean volumes derived from 3-D reconstructed MR images for the two bony fragments were 10,363.26 mm 3 and 8273.27 mm 3 (percentage measurement error (PME) of 2.21% and 2.49%, respectively). The mean fragment separation was 2.97 mm (PME 3.98%). The mean raw measurement error when measuring the degree of fragment depression was 0.31 mm and this was considered clinically acceptable. Conclusion The findings of this study suggest that 3-D reconstructed MR imaging may be sufficiently accurate for the evaluation of tibial plateau depression, fracture gap and fragment volume.