Quantitative Study Of Early Cervical Spinal Cord Compression By DTI And PSIR

Objective: In this study,phase-sensitive inversion recovery sequence(PSIR)combined with zoomed diffusion tensor imaging(ZOOM-DTI)was used to quantitatively evaluate patients with early cervical spinal cord compression in order to obtain effective indicators for early diagnosis and disease monitoring.Materials and methods: Sixty-two patients with early cervical spinal cord compression(16 males,46 females;Median age 40(27.5～54)years,range 22 to 78years)were recruited in this study.Cervical spinal segments were divided into compression group and non-compression group according to whether they were compressed or not.They were divided into symptomatic group(30 cases)and asymptomatic group(32 cases)according to the presence and absence of clinical symptoms.PSIR and DTI were acquired on a 3.0T MR scanner(Ingenia CX,Philips Healthcare,the Netherlands)with a 32-channel head coil.The image quality of the PSIR imaging was evaluated by 3 grades.The region of interest(ROIS)was manually drawn on the fusion image of cervical spinal cord DTI and PSIR,and placed on the bilateral anterior horn of gray matter and bilateral anterior,middle and posterior columns of white matter in each segment.The fractional anisotropy(FA),apparent diffusion coefficient(ADC),axial dispersion coefficient(AD)and relative anisotropy index(RA)were recorded.In order to ensure the accuracy of the measured data and the repeatability of this study,20 patients were randomly selected for the second measurement,and the interval of the two measurements was two months.Intraclass correlation coefficient(ICC)was used to compare the consistency of the two measurements.Pearson correlation test was used to test the correlation between DTI parameters and age in the uncompresssed group.One-way ANOVA was used to compare the DTI parameters of gray matter and white matter columns in different sections of uncompressed cervical spinal cord.Wilkerson symbolic rank test was used to compare the gender,lateral and ipsilateral differences of DTI parameters in the unpressured group,and the differences between the compression group and the uncompresssed group as well as the asymptomatic group and the symptomatic group.The chi-square test was used to compare and analyze the basic clinical data of patients with and without symptoms.Binary logistics regression analysis and receiver operating characteristic(ROC)were used to analyze the diagnostic efficacy of different DTI parameters and the joint variables between symptomatic,asymptomatic and noncompression groups.Results:1 PSIR images of 10 patients were grade 0 and the imaging of 52 patients was grade 1.All meet the needs of diagnosis and measurement.2 The ICC values obtained from the two measurements were all greater than 0.75,indicating that the measurement of DTI parameters in this study has good consistency and repeatability.3 In non-compression group,there was no gender difference in the DTI parameters of cervical spinal cord(P > 0.05).Some DTI parameters were correlated with age,including FA values of the left gray matter(r=-0.298,P=0.027),left posterior column(r=-0.371,P=0.005),right posterior column(r=-0.344,P=0.010)and RA values of the left gray matter(r=-0.360,P=0.018),and left posterior column(r=-0.412,P=0.006)of each cervical spinal cord segment.Comparison of columns of gray matter and white matter in upper,middle and lower segments,the AD values of the left middle column(P=0.009),the left posterior column(P=0.026),the right gray matter(P=0.000),the right anterior column(P=0.025),the FA values of the left gray matter(P=0.015),the left posterior column(P=0.024),the right posterior column(P=0.029),and the RA values of the left gray matter(P=0.002),the left middle column(P=0.043),and the left posterior column(P=0.019)were significant differences and decrease gradually from top to bottom.There were also statistically significant differences in DTI parameters of bilateral gray and white matter and gray and white matter in the same segment(P<0.05).Compared with the contralateral,FA values of the right gray matter and left white matter and RA values of the right anterior column were increased,with statistical significance(P<0.05).The ADC and AD values in the right gray matter,the AD values in the right anterior column and the left middle column decreased,with statistical significance(P<0.05).Compared with the gray matter on the same side,ADC,RA,AD and FA in the white matter on both sides were increased,while FA and RA values in the anterior column on both sides were decreased,with statistical significance(P<0.05).4 In compression group,compared with non-compression group,FA value of right gray matter decreased,FA value of right anterior column increased(P<0.05),ADC value of left anterior column and right white matter decreased(P<0.05),RA value and AD value of right gray matter decreased(P<0.05),AD value of right anterior column decreased(P<0.05).5 In symptomatic group,the age was greater than the asymptomatic group(P=0.001)while there was no difference in gender(P>0.05).Increased AD in right white matter(P<0.05),increased FA in right anterior column(P<0.05),increased ADC in bilateral white matter(P<0.05),increased FA in right gray matter and decreased FA in bilateral anterior column was found(P<0.05).Among them,the AD values of the right middle column set 2.150 as the threshold values for the diagnosis of cervical spinal cord compression,the area under the ROC curve(AUC)was 0.686,the diagnostic sensitivity was 80%,the specificity was 55%,and the 95% confidence intervals was(0.519,0.854).Combined variables were used to diagnose cervical spinal cord compression,the threshold value was 0.641 for the diagnosis of cervical spinal cord compression,and the AUC was 0.830.The sensitivity and specificity were 65.0% and100.0%,respectively,and the 95% confidence intervals were(0.693,0.967).In asymptomatic group,increased AD in left gray matter(P<0.05);decreased FA in right middle column(P<0.05);increased RA in right gray matter was found(P<0.05).The area under the ROC curve(AUC)was 0.666,the sensitivity was 72%,the specificity was 56%,and the 95% confidence interval was(0.514,0.817),when the right gray matter RA value was 0.550.Setting the left gray matter AD value of 1.45 as the threshold for the diagnosis of cervical spinal cord compression,AUC was 0.666,the diagnostic sensitivity was 95.7%,the specificity was 26.1%,and the 95% confidence intervals were(0.510,0.823).Combined variables were used to diagnose cervical spinal cord compression,the threshold value was 0.391 and the AUC was 0.744.The diagnostic sensitivity and specificity were 78.3% and 60.9%,respectively,and the 95% confidence intervals were(0.603,0.844).As for the comparation between the asymptomatic patients group and symptomatic patients group,symptomatic patients group decreased AD was found in right posterior column(P<0.05);decreased ADC was found in bilateral posterior column(P<0.05);decreased RA was found in left middle column(P<0.05).The AD value of the right posterior column was used as the threshold to diagnose cervical spinal cord compression with symptoms,AUC was 0.690,the diagnostic sensitivity was 47.8%,the specificity was 85%,and the 95% confidence interval was(0.532,0.849).Combined variables were used to diagnose cervical spinal cord compression,the threshold value was 0.479,the AUC was 0.814,the sensitivity was 75%,the specificity was 60.9%,and the confidence interval was 78.3%(0.688,0.940).Conclusion:1.PSIR can clearly display the subanatomy of cervical spinal cord gray and white matter.ZOOM-DTI improves the accuracy of measurement of the dispersion state of gray and white matter,and the combination of the two can measure the substructure of cervical spinal cord more accurately.2.There was no gender difference in the DTI parameters of uncompressed cervical spinal cord,and age and segmental dependence existed.In the same segment,DTI parameters of gray matter and white matter showed lateral differences,as well as differences between gray matter and white matter.3.The quantitative analysis of ZOOM-DTI combined with PSIR can show the microstructural changes of gray and white matter caused by long-term cervical spinal cord compression before the appearance of clinical symptoms,especially the white matter.The combination of variables will further improve the diagnostic capability.