Study On Mild Cervical Spondylotic Myelopathy Based On IVIM And MRI Texture Analysis

PART1Diagnostic value of incoherent motion diffusion-weighted imaging in mild cervical spondylotic myelopathyObjective:To investigate the value of intravoxel incoherent motion diffusion-weighted imaging(IVIM-DWI)in the diagnosis of mild cervical spondylotic myelopathy(CSM).Methods:A total of 60 patients with CSM who were clinically diagnosed with CSM from November 2021 to June 2022 in Yijishan Hospital of Wannan Medical College were collected.They were divided into mild group(40 cases,13-16 points,no high signal on T2WI spinal cord compression),moderate and severe group(20 cases,5-12 points,T2WI with high signal on spinal cord compression),and another 40 cases were selected.Healthy adult volunteers served as the control group.All subjects underwent cervical spine conventional MRI and IVIM multi-b-value scan.The values of apparent diffusion coefficient(ADC),true diffusion coefficient(D),false diffusion coefficient(D*)and perfusion fraction(f)were measured respectively.Spearman correlation was used to analyze the correlation between each parameter value and JOA score,one-way analysis of variance was used to compare the differences of each parameter among the groups,and receiver operating characteristic(ROC)curve was drawn to evaluate the efficacy of each quantitative parameter and its combination in diagnosing CSM.P<0.05 was considered statistically significant.Results:In the control group,mild group and moderate-severe group,D value and ADC value showed a gradually increasing trend,while D*value and f value showed a gradually decreasing trend.There were significant differences in D value,D*value,f value and ADC value among the three groups(all P<0.01),and there were also statistical significance in pairwise comparisons(all P<0.01).JOA score was negatively correlated with D value(r=-0.680,p<0.01),positively correlated with D*value(r=0.489,p<0.01),and positively correlated with f value(r=0.707,p<0.01),It was negatively correlated with ADC value(r=-0.712,p<0.01).The diagnostic efficacy of D value and ADC value in the control group and mild group was higher(AUC was 0.925,0.918,respectively),and the combined diagnostic efficacy of each parameter was the highest(AUC was 0.999).When the critical value was 0.975,its sensitivity and specificity The accuracy and accuracy are 1.000,0.975,and 0.987,respectively.The diagnostic efficacy of D value and ADC value in the mild group and moderate-severe group were also high(AUC were 0.716,0.866,respectively),and the combined diagnostic efficacy of each parameter was the highest(AUC was 0.985).When the critical value was 0.900,its sensitivity,specificity and accuracy were 0.950,0.950,and 0.950,respectively.Conclusions:The intravoxel incoherent motion diffusion-weighted imaging technique can quantitatively evaluate CSM,detect CSM spinal cord injury early,and show its severity.PART 2Study on mild cervical spondylotic myelopathy based on conventional MRI texture analysis techniqueObjective:To explore the value of conventional MRI texture analysis in the study of mild cervical spondylotic myelopathy.Methods:A total of 60 patients with CSM diagnosed clinically in Yijishan Hospital of Southern Anhui Medical College from November 2021 to June 2022 were divided into mild group(40 cases,13:16,no high signal intensity in T2WI spinal cord compression site)and moderate severity group(20 cases,5-12 points,T2WI spinal cord compression site showed high signal)according to the evaluation standard of spinal cord injury function of Japanese Orthopaedic Association(JOA score)and the difference of spinal cord signal on MRI-T2WI image.Another 40 healthy adult volunteers were selected as the control group.All subjects underwent routine MRI scan of cervical vertebrae.The T2WI images of all the subjects were imported into the ITK-SNAP software.According to the compressed segments of the spinal cord in the CSM group,the ROI was manually sketched layer by layer along the four layers of the spinal cord C3/4-C6/7 on the T2 sagittal images,and fused into a three-dimensional volume region of interest,then the texture features were extracted by AK software,and the texture features with statistically significant differences between groups were screened out by independent sample t-test or Mann-WhitneyU test.After that,the optimal texture feature parameters are further screened by multi-factor logistic regression,and the texture feature model is established.The receiver operating feature(ROC)curve was drawn to evaluate the efficacy of each texture feature parameter and its joint diagnosis of CSM.Results:A total of 107 texture feature parameters are extracted based on MRI-T2WI.In the control group and the mild group,four optimal texture parameters were selected,which were glcm_Imc1,gldm_DependenceNonUniformity,shape_Elongation and shape_Sphericity,in which the glcm_Imc1 value of the mild group was higher than that of the control group,and the other three texture parameters were lower than those of the control group.Finally,an optimal texture parameter,glcm_Correlation,was screened out in the mild group and the moderate to severe group,which was lower than that in the moderate and severe group,and the difference was statistically significant.The diagnostic efficacy of the parameters of shape_Elongation and shape_Sphericity texture features was higher in the control group and mild group(AUC=0.709,0.804,respectively),while the diagnostic efficiency of the combined texture feature model was the highest(AUC=0.892).When the critical value was 0.623,its sensitivity,specificity,accuracy,positive predictive value and negative predictive value were 0.900,0.775,0.838,0.800 and 0.866,respectively.The AUC of glcm_Correlation characteristic parameters in mild group and moderate to severe group was 0.716.When the critical value was 0.725,the sensitivity,specificity,accuracy,PPV and NPV were 0.900,0.825,0.850,0.720 and 0.943,respectively.Conclusions:Conventional MRI texture analysis technology has a certain diagnostic efficiency for early cervical Spondylotic myelopathy,can transform image visual information into deep-seated features,and provide technical support for clinical diagnosis and intervention treatment.