Preoperative liver volumetry: how does the slice thickness influence the multidetector computed tomography- and magnetic resonance-liver volume measurements?

Abstract

OBJECTIVE: The purpose was to investigate the influence of slice thickness on multidetector computed tomography (MDCT)- and magnetic resonance (MR)-based liver volumetry. MATERIALS AND METHODS: Twenty patients who underwent liver surgery were imaged with either a 64-slice MDCT (n = 10) or a 1.5-T MR scanner (n = 10). Multidetector computed tomography and MR images were reconstructed using different slice thicknesses (2, 4, 6, and 8 mm). Total liver volumes (TLVs) were measured by 2 independent readers based on different slice thicknesses using semiautomatic software. Results were compared with TLVs based on 2-mm slices that served as standard of reference. The time to perform each volumetry was recorded. RESULTS: For MDCT volumetry, a statistical difference was seen only between TLVs based on 2-mm versus 8-mm slices (P = 0.012 and P = 0.002 for readers 1 and 2, respectively). For MR volumetry, no statistical difference was seen between TLVs of the standard of reference and TLVs based on 4-, 6-, and 8-mm slices. Regarding the time to perform volumetry, there was a significant gain of time for both readers when volumetry was performed on 6- and 8-mm MDCT slices and on 4-, 6-, and 8-mm MR slices (P < 0.0167) when compared with the standard of reference. CONCLUSIONS: The results of MDCT- and MR-based liver volumetry are dependent on slice thickness. With respect to the precision of calculated volumes and the significant gain of time, 6-mm slices are preferable for computed tomographic imaging, and 8-mm slices are preferable for MR imaging.