Quantitative estimation of the high-intensity zone in the lumbar spine: comparison between the symptomatic and asymptomatic population

BACKGROUND CONTEXT: The high-intensity zone (HIZ) on magnetic resonance imaging (MRI) has been studied for more than 20 years, but its diagnostic value in low back pain (LBP) is limited by the high incidence in asymptomatic subjects. Little effort has been made to improve the objective assessment of HIZ. PURPOSE: To develop quantitative measurements for HIZ and estimate intra- and interobserver reliability and to clarify different signal intensity of HIZ in patients with or without LBP. STUDY DESIGN: A measurement reliability and prospective comparative study. PATIENT SAMPLE: A consecutive series of patients with LBP between June 2010 and May 2011 (group A) and a successive series of asymptomatic controls during the same period (group B). OUTCOME MEASURES: Incidence of HIZ; quantitative measures, including area of disc, area and signal intensity of HIZ, and magnetic resonance imaging index; and intraclass correlation coefficients (ICCs) for intra-and interobserver reliability. METHODS: On the basis of HIZ criteria, a series of quantitative dimension and signal intensity measures was developed for assessing HIZ. Two experienced spine surgeons traced the region of interest twice within 4 weeks for assessment of the intra-and interobserver reliability. The quantitative variables were compared between groups A and B. RESULTS: There were 72 patients with LBP and 79 asymptomatic controls enrolling in this study. The prevalence of HIZ in group A and group B was 45.8% and 20.2%, respectively. The intraobserver agreement was excellent for the quantitative measures (ICC=0.838-0.977) as well as interobserver reliability (ICC=0.809-0.935). The mean signal of HIZ in group A was significantly brighter than in group B (57.55 +/- 14.04% vs. 45.61 +/- 7.22%, p=.000). There was no statistical difference of area of disc and HIZ between the two groups. The magnetic resonance imaging index was found to be higher in group A when compared with group B (3.94 +/- 1.71 vs. 3.06 +/- 1.50), but with a p value of .050. CONCLUSIONS: A series of quantitative measurements for HIZ was established and demonstrated excellent intra-and interobserver reliability. The signal intensity of HIZ was different in patients with or without LBP, and significant brighter signal was observed in symptomatic subjects. (C) 2014 Elsevier Inc. All rights reserved.