Three-dimensional reconstructed MRI of an acrylic meniscal cartilage phantom: The effect of acquisition slice thickness upon accuracy of volume measurement

3. Dr. Karen J. Reynolds School of Informatics and Engineering, Flinders University, GPO Box 2100, Adelaide, South Australia 5001 Tel: +61 8 8201 5190 Fax: +61 8 8201 3618 e-mail: [email protected] ABSTRACT The aim of this study was to determine the influence of slice thickness upon the accuracy of volume measurement taken from three-dimensional reconstructed magnetic resonance (MR) images. A pair of scaled meniscal cartilage phantoms (medial and lateral) were created from a low viscosity acrylic casting resin. The two phantoms were imaged simultaneously using a modified proton density turbo spin-echo (PD TSE) MR sequence at slice thicknesses of 4.0, 3.0, 2.0, 1.5 and 1.0mm. At each slice thickness, a three-dimensional reconstructed image of each phantom was created, using a commercially available software package. Using the software’s inbuilt analysis functions, a volume measurement was then made of each image. The software generated volume measures were compared to a “gold standard” mass/density measure of the phantoms. Percentage measurement error (PME) measures were calculated, with mean values ranging from 19.42% (4.0 mm slice thickness) to 1.05% (1.0 mm). Paired t testing suggested no statistically significant difference (P = 0.15) between the means of the PME values for the two phantoms suggesting an acceptable degree of measurement consistency. In conclusion, the results of this study suggest that the three-dimensional reconstructed technique can be used to achieve a highly accurate volume measurement (PME < 5%) of irregular (non-geometric) objects or image components (such as human meniscal cartilages) without the need to employ sub-millimetre acquisitions. Such slice thicknesses are within the capabilities of most commercial MR scanners, making the results reported here relevant to the majority of clinical imaging sites.