Comparison of methods to assess quadriceps muscle volume using magnetic resonance imaging

• Published in: Journal of magnetic resonance imaging Vol. 30; no. 5; pp. 1116 - 1123
• Main Authors: Nordez, Antoine, Jolivet, Erwan, Südhoff, Ingrid, Bonneau, Dominique, de Guise, Jacques A., Skalli, Wafa
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc., A Wiley Company 01.11.2009; Wiley-Blackwell
• Subjects: Adult; Cavalieri formula; Humans; Image Processing, Computer-Assisted - methods; Imaging, Three-Dimensional - methods; interpolation; Leg; Life Sciences; Magnetic Resonance Imaging - methods; Male; Models, Biological; Models, Statistical; muscle cross sectional area; Muscle, Skeletal - pathology; Quadriceps Muscle - pathology; Reproducibility of Results; truncated cone
• ISSN: 1053-1807; 1522-2586; 1522-2586
• DOI: 10.1002/jmri.21867

Purpose To compare the precision of four methods to estimate the volume of quadriceps muscles using axial MRI. Materials and Methods Entire legs of 10 healthy young subjects were scanned using a 1.5 Tesla magnetic resonance imaging scanner and 4‐mm‐thick sections without any gaps. Quadriceps muscles were outlined on all of the slices to obtain the MRI reference standard measure of quadriceps muscle volume. This MRI reference standard was compared with the volume estimated using (i) the truncated cone formula, (ii) the Cavalieri method, (iii) a cubic spline interpolation of missing cross sectional areas, and, (iv) the deformation of a parametric specific object. For each method, 3 to 21 slices were used. Results The average volume error was significantly (P < 0.001) different in comparing the four methods (4.4%, 2.3%, 1.1%, and 1.2%, respectively). In addition, the number of slices required to reach a given volume error was significantly (P < 0.001) different across all methods (respectively, 12, 9, 5, and 7 slices required to reach a volume error of 1.1%). Conclusion While methods based on interpolation and deformation of a parametric specific object have not been used in literature, these two methods are the most precise approaches to reach a given level of precision. J. Magn. Reson. Imaging 2009;30:1116–1123. © 2009 Wiley‐Liss, Inc.