Clinical correlation of division measurement of dural sac in MRI with lumbar spinal central stenosis

Bo WANG,Chen WANG,Yun-tao WANG,Yun-feng RUI,Hui CHEN
DOI: https://doi.org/10.3877/cma.j.issn.1674-0785.2013.24.036
2013-01-01
Abstract:Objective In MRI imaging, the dural sac were divided into three different regions, and study their clinical relevance with lumbar spinal central stenosis(LSCS). Methods Oswestry disability index (ODI), visual analog scale (VAS), walking distance and MRI were completed voluntarily by all these 126 patients. The whole dural cross-sectional area were measured in MRI T2 weighted transaxial images. Then, the dural sac was divided into three different regions, which were defined as left, right and middle area and measured them respectively. Intra- and interobserver reliability were assessed by ANOVA (P value). Spearman correlation coefficients(rs value) were used to assess the correlations between the areas and LSCS baseline characteristics. The correlations between the areas and ODI, VAS or walking distance were investigated by Pearson correlation coefficient (r value). Results Intra- and interobserver reliability had no statistical significance (P>0.05). A negative appearance was identified in the whole and middle dural cross-sectional area (P>0.05). The correlations between the areas and LSCS baseline characteristics had no statistical significance (P>0.05). The left region of dural sac in MRI had relationship with VAS of back pain (r=-0.397) and ODI (r=-0.404). The right region of dural sac in MRI correlated with VAS (back pain: r=-0.422, leg pain: r=-0.392), ODI (r=-0.443) and walking distance (r=-0.393), especially when either area was lower than 25 mm2, the correlations with ODI were more significantly (r=-0.621 and -0.635, respectively, P<0.01). Conclusion Region division can improve correlation of radiological methods with LSCS clinical symptoms, while the whole region of dural sac could not detect. The left and right region of dural sac in MRI T2 weighted transaxial images significantly correlated with the&nbsp;severity of LSCS symptoms.
What problem does this paper attempt to address?