Reliability and agreement of the diagnosis criteria for cervical intervertebral disc injury based on the manifestations of X-Ray, CT and MRI examinations
Enchao GUAN,Shengdan JIANG,Yuren WANG,Bo LI,Xinfeng ZHENG,Leisheng JIANG
DOI: https://doi.org/10.3969/j.issn.2095-9958.2017.05-01
2017-01-01
Abstract:Background:The hyperextension injury of the cervical spine is a common injury, which usually leads to cervi-cal intervertebral disc rupture and may result in cervical instability and other secondary lesions. Early, timely and accurate diagnosis will contribute to good clinical outcomes of treatment and recovery. Objective: To evaluate the reliability and agreement of the characteristic findings on X-Ray, CT and MRI examinations for the diagnosis of cervical intervertebral disc injury. Methods:A total of 44 patients who had hyperextension injury of the cervical spine and underwent anterior ap-proach operations between November 2013 and July 2016 were included in this study. Each of these patients had complete documentations of preoperative images including X-ray, CT and MRI, but no apparent evidence of vertebral fractures or dis-locations. Characteristic findings were defined as diagnostic criteria for cervical intervertebral disc injury. The imaging doc-umentations were read by three physicians separately, and the diagnosis was compared with intraoperative findings. Inter-ob-server agreement for each assessment was determined using the Kappa statistic. The sensitivity, specificity, accuracy, posi-tive and negative predictive values of different image features in detecting cervical disc injury were calculated. The intraop-erative findings were used as the gold standard. Results:The reliability and inter-observer agreement of each diagnostic cri-terion in X-Ray and CT for the assessment of cervical disc injury were lower than those in MRI. Thickening of the soft tis-sue anterior to the vertebral body had both high specificity (89.6%-94.8%) and accuracy (80.6%-83.9%) in assessment of the cervical disc injury. Inter-observer agreement of discontinuity on sagittal T1–weighted image (T1D) for the assessment of cervical disc injury was fair (Kappa value was 0.217, 0.267 and 0.318) , and its sensitivity (43.8%-63.5%) and positive predictive value (25.9%-33.3%) were relatively low. For longitudinal high signal intensity on sagittal T2–weighted image (T2L), the inter-observer agreement was moderate (Kappa value was 0.490, 0.439 and 0.420), and its sensitivity (68.8%-75.0%) and positive predictive value (35.5%-42.4%) were higher than those in T1D. However, the inter-observer agreement of transversal high or intermediate signal intensity on sagittal T2-weighted image (T2T) for the assessment of cervical disc injury was good (Kappa value was 0.669, 0.656 and 0.614), and its sensitivity (81.3%-87.5%), specificity (93.5%-96.1%), accuracy (92.4%-93.5%), positive predictive value (73.7%-81.3%) and negative predictive value (96.0%-97.3%) were all high. Conclusions:When thickening of the soft tissue anterior to the vertebral body is noted in X-Ray and CT images, cervi-cal disc injury may be highly suspected. T2T of MRI is a reliable indicator for cervical disc injury. MRI seems to be more predictable than X-Ray and CT in excluding the cervical disc injury, but this advantage was not statistically significant.