• Ning Zhang, Xianjun Zeng, Laichang He, Zhili Liu, Jiaming Liu, Zhihong Zhang, Xuanyin Chen, and Yong Shu.
• Department of Radiology, The First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, China.
• World Neurosurg. 2019 Aug 1; 128: e806-e813.

ObjectiveDifferentiating pyogenic spondylitis (PS) and tuberculous spondylitis (TS) is challenging but critical in clinical practice. Accurate diagnosis and early treatment are crucial to preventing further progression of disease. Magnetic resonance imaging (MRI) is considered the best method for the diagnosis of spinal infection, but results remain imprecise; therefore, by comparing and analyzing the MRI findings of spinal infections, we intend to identify key distinguishing features between PS and TS, and with that establish a systematic scoring method to help clinicians.MethodsThe MRI features of 70 cases of spinal infection (32 cases of PS, 38 cases of TS), confirmed by pathology or clinical diagnosis, were retrospectively analyzed. The receiver operating characteristic curve was used to analyze the diagnostic efficacy of the MRI parameter scores in PS and TS.ResultsAmong the 70 cases, the average age was 54.5 years, and 43 were male. Sixteen parameters were significantly different between the PS and TS groups. We hypothesized that a diagnosis of PS could be made when the number of parameters characteristic of PS exceeded the number of parameters characteristic of TS, and vice versa. We randomly selected 70% (49 patients) of the 70 patients for analysis, and then validated in the remaining 30% (21 patients) of cases. Using 0.5 as the cutoff value, of the remaining 21 patients, the correct coincidence rate was 95.23%, sensitivity was 91.67%, specificity was 100%, false-positive rate was 0%, and false-negative rate was 8.3%. The MRI parameter scores of PS and TS were analyzed with the receiver operating characteristic; area under the curve was 1.00.ConclusionsOur systematic scoring system of MRI parameters is helpful in differentiating PS and TS.Copyright © 2019 Elsevier Inc. All rights reserved.

27 keywords...

hide…