To investigate the predictive value of preoperative conventional MR-T2WI histograms on texture of spinal meningioma.

The clinical data of 40 patients with spinal meningioma admitted to the Neurosurgery Department of 900^th Hospital of Joint Logistics Team of People's Liberation Army from January 2015 to May 2022 were analyzed retrospectively. The tumors were divided into soft and rigid groups according to the difficulty of intraoperative resection. All patients underwent 3.0T MRI scan before surgery, selected the maximum diameter level of the tumor in the T2WI sagittal position by ImageJ software and delineated the region of interest (ROI), and measured the histogram related parameters of ROI, including the mean, median, crowd, maximum, minimum, standard deviation, bias, and kurtosis. The differences between the parameters in the soft and hard texture groups were compared, and the receiver operating characteristics (ROC) curves were drawn between the statistically significant parameters, and the predictive efficacy of T2WI histogram on spinal meningioma texture was analyzed.

The maximum, minimum, standard deviation, bias, and kurtosis value of T2WI histograms in 2 groups were not significantly different (P>0.05). The mean, median, and medium values of the T2WI histogram signals in the soft texture group were significantly larger than those in the hard texture group, and the difference was statistically significant (P<0.05). The area under the ROC curve for the mean, median, and crowd of the tumor soft texture group was 0.801, 0.804, and 0.792, respectively. The cut-off value of the mean was 272.290, the discrimination sensitivity was 91.67% and the specificity was 60.70%. The median cut-off value was 373.680, the discrimination sensitivity was 66.70% and the specificity was 85.70%. The crowd cut-off value was 295.350, the discrimination sensitivity was 83.30% and the specificity was 67.90%.

The mean, median, and medium number of the T2WI histogram signals can be used to predict the texture of the spinal meningioma of the tumor, among which the median is better.