Reliability of the Goutallier classification in quantifying muscle fatty degeneration in the lumbar multifidus using magnetic resonance imaging

Objective: The purpose of this study was to investigate the reliability of the Goutallier classification system (GCS) for grading muscle fatty degeneration in the lumbar multifidus (LM) using magnetic resonance imaging (MRI) examinations.

Methods: Lumbar spine MRI scans were obtained retrospectively from the radiology department imaging system. Two examiners (a chiropractic diagnostic imaging resident and a board certified chiropractic radiologist with 30 years of experience) independently graded each LM at the L4/5 and L5/S1 intervertebral level. ImageJ pixel analysis software (version 1.47; National Institutes of Health, Bethesda, MD) was used independently by 2 observers to quantify the percent fat of the LM and allow correlation between LM percent fat and GCS grade. Twenty-five subject MRIs were randomly selected. Magnetic resonance imaging scans were included if they were obtained using a 1.5 T imaging system and were excluded if there was evidence of spinal infection, tumor, fracture, or postoperative changes. For all tests, P < .05 was defined as significant.

Results: Intraobserver reliability grading LM fat ranged from a weighted κ (κw) of 0.71 to 0.93. Mean interobserver reliability grading LM fat was κ(w), 0.76 to κ(w), 0.85. There was a significant (P < .001) correlation between LM percent fat and GCS grade. Furthermore, interobserver reliability in determining percent fat was between intraclass correlation coefficient, 0.73 to intraclass correlation coefficient, 0.90.

Conclusions: In this study, the GCS was reliable in grading LM fatty degeneration and correlated positively with a quantified percent fat value. In addition, ImageJ software (National Institutes of Health) was reliable between raters when quantifying LM percent fat.