Effect of Number of Measurement Points on Accuracy of Muscle T2 Calculations.
10.13104/imri.2016.20.4.207
- Author:
Noriyuki TAWARA
1
;
Atsushi NISHIYAMA
Author Information
1. Department of Radiological Sciences, Faculty of Health Sciences, Japan Health Care College, Sapporo, Japan. n_tawara@nihoniryo-c.ac.jp
- Publication Type:Original Article
- Keywords:
Transverse relaxation time;
Skeletal muscle;
Number of measurement points;
Goodness-of-fit;
T2 effective signal-ratio
- MeSH:
Humans;
Magnetic Resonance Imaging;
Male;
Muscle, Skeletal;
Polyvinyl Alcohol;
Relaxation;
Volunteers
- From:Investigative Magnetic Resonance Imaging
2016;20(4):207-214
- CountryRepublic of Korea
- Language:English
-
Abstract:
PURPOSE: The purpose of this study was to investigate the effect of the number of measurement points on the calculation of transverse relaxation time (T2) with a focus on muscle T2. MATERIALS AND METHODS: This study assumed that muscle T2 was comprised of a single component. Two phantom types were measured, 1 each for long (“phantom”) and short T2 (“polyvinyl alcohol gel”). Right calf muscle T2 measurements were conducted in 9 healthy male volunteers using multiple-spin-echo magnetic resonance imaging. For phantoms and muscle (medial gastrocnemius), 5 regions of interests were selected. All region of interest values were expressed as the mean ± standard deviation. The T2 effective signal-ratio characteristics were used as an index to evaluate the magnetic resonance image quality for the calculation of T2 from T2-weighted images. The T2 accuracy was evaluated to determine the T2 reproducibility and the goodness-of-fit from the probability Q. RESULTS: For the phantom and polyvinyl alcohol gel, the standard deviation of the magnetic resonance image signal at each echo time was narrow and mono-exponential, which caused large variations in the muscle T2 decay curves. The T2 effective signal-ratio change varied with T2, with the greatest decreases apparent for a short T2. There were no significant differences in T2 reproducibility when > 3 measurement points were used. There were no significant differences in goodness-of-fit when > 6 measurement points were used. Although the measurement point evaluations were stable when > 3 measurement points were used, calculation of T2 using 4 measurement points had the highest accuracy according to the goodness-of-fit. Even if the number of measurement points was increased, there was little improvement in the probability Q. CONCLUSION: Four measurement points gave excellent reproducibility and goodness-of-fit when muscle T2 was considered mono-exponential.
