Ultrasound and MRI measured changes in muscle mass gives different estimates but similar conclusions: a Bayesian approach

Eur J Clin Nutr. 2019 Aug;73(8):1203-1205. doi: 10.1038/s41430-019-0431-z. Epub 2019 Apr 23.

Abstract

The purpose of this paper was to use a Bayesian approach to compare the relative change in muscle size between magnetic resonance imaging (MRI) and ultrasound measured muscle thickness (MTH) following 6 weeks of concentric and eccentric blood flow restricted exercise. Changes at each site were as follows: concentric 50% site (MRI: 10.2%, MTH: 8.7%), concentric 10 cm site (MRI: 12%, MTH: 4.5%), eccentric 50% site (MRI: -1.7%, MTH: 2.6%), and eccentric 10 cm site (MRI: 5.2%, MTH: 0.5%). When testing the difference between estimates using a default prior of 0.707, we provided evidence that the estimate at the 50% site of the concentric arm was similar between ultrasound and MRI [Median % (95% credible interval): -1.1 (-8.2, 5.8)]. However, evidence for other sites suggested differences or a degree of uncertainty. Both methods produce similar conclusions about the presence of growth but the magnitude of that change appears different at most sites.

MeSH terms

  • Bayes Theorem
  • Exercise / physiology*
  • Humans
  • Magnetic Resonance Imaging / methods*
  • Muscle, Skeletal / anatomy & histology*
  • Muscle, Skeletal / physiology*
  • Reproducibility of Results
  • Ultrasonography / methods*