The Impact of Slice Interval and Equation on the Accuracy of Magnetic Resonance Image Estimation of Quadriceps Muscle Volume in End Stage Liver Disease

Abstract

Introduction: End stage liver disease (ESLD) is associated with loss of muscle mass and function, known as sarcopenia, which can increase the risk of complications of ESLD, hospitalization and mortality. Therefore, the accurate assessment of muscle mass is essential to evaluate sarcopenia in ESLD. However, manual segmentation of muscle volume (MV) can be laborious on cross-sectional imaging, due to the number of slices that require analysis. This study aimed to investigate the impact of reducing the number of slices required for MV estimation. Further, we aimed to compare two equations utilized in estimating MV (cylindrical and truncated cone). Methods: Thirty eight ESLD patients (23 males; 54.8 ± 10.7 years) were recruited from the Queen Elizabeth University Hospital Birmingham. A 3T MRI scan was completed of the lower limbs. Quadriceps MV was estimated utilizing 1-, 2-, 3-, and 4 cm slice intervals with both cylindrical and truncated cone equations. Absolute and relative error (compared to 1 cm slice interval) was generated for 2-, 3-, and 4 cm slice intervals. L3 skeletal muscle index (SMI) was also calculated in 30 patients. Results: Relative error increased with slice interval using the cylindrical (0.45 vs. 1.06 vs. 1.72%) and truncated cone equation (0.27 vs. 0.58 vs. 0.74%) for 2, 3, and 4 cm, respectively. Significantly, the cylindrical equation produced approximately twice the error compared to truncated cone, with 3 cm (0.58 vs. 1.06%, P < 0.01) and 4 cm intervals (0.74 vs. 1.72%, P < 0.001). Finally, quadriceps MV was significantly correlated to L3 SMI (r2 = 0.44, P < 0.0001). Conclusion: The use of the truncated equation with a 4 cm slice interval on MRI offers an efficient but accurate estimation of quadricep muscle volume in ESLD patients.