Reliability of histographic analysis for paraspinal muscle degeneration in patients with unilateral back pain using magnetic resonance imaging.

Abstract

Mixed use of T1- or T2-weighted magnetic resonance imaging (MRI) in histographic analysis has impeded cross-reference of studies on fatty degeneration of muscles in patients with back pain. The current study aimed to establish the measurement reliability of histographic analysis for normal and painful paraspinal muscles between T1- and T2-weighted MRI. T1- and T2-weighted MRI at the L4-5 level were performed in 31 patients with unilateral back pain. Two raters conducted the histographic analysis on 2 separate occasions. Multifidus, erector spinae, and psoas major muscles were assessed on both normal and painful sides. Intra- and interrater reliability of the parameters were generally excellent for both T1- (intrarater: 0.83 ± 0.13 and interrater: 0.93 ± 0.08) and T2-weighted images (intrarater: 0.93 ± 0.12 and interrater: 0.92 ± 0.12). Histogram width was fair to good for both kinds of image. Intraclass correlation coefficients were similar between painful and normal sides. The evidence that most histographic analysis parameters had excellent reliability for both T1- and T2-weighted images allows for cross-reference of the results between the two kinds of image. Moreover, the results of this study support the availability of histographic analysis for assessment of fatty degeneration in patients with back pain.