| Part I Diffusion tensor imaging of tibial nerve and commonperoneal nerve in normal volunteersObjectiveDiffusion tensor imaging(DTI)is a new functional magnetic resonance imaging technique,which is based on diffusion weighted imaging(DWI).It can display the microstructure of tissue noninvasively.The propose of our study is to explore the diffusion characteristics of water molecules in tibial nerve(TN)and common peroneal nerve(CPN)tissue through fiber track and the measurement of fractional anisotropy(FA)and apparent diffusion coefficient(ADC),then to explore the appropriate b-value in our study.Materials and Methods1.Participant DataTwelve healthy adult volunteers(4 men.8 women;mean age 49 years,range 28-65 years)were enrolled from May to December 2015.Admission criteria were no general contraindications for magnetic resonance imaging(MRI),no pregnancy,no history of leg or knee surgery.no history of exposure to neurotoxic agents and cardiovascular.pulmonary,endocrince neurological.neuromuscular or no knee related musculoskeletal disorders,no other chronic diseases.2.MR protocolAll volunteers were performed bilateral knee 3.0T MR(Achieva Philips,Healthcare,Best,Netherlands)scanning.An eight-channel knee coil was employed.The subjects were placed supine with feet-first position and unilateral imaging(first right and then left),then a total of 24 knee images were obtained.MRI sequences included two-dimensional(2D)axial T1-weighted imaging(T1WI).2D axial T2-weighted spectral adiabatic inversion recovery imaging(SPAIR T2WI)and DTI.The imaging parameters were as follows.T1WI:repetition time/echo time(TR/TE)550/20 ms,slice thickness 3.0 mm,overlap 0.3 mm,field of view(FOV)130 ×130 mm2,number of slices 45,sensitivity encoding factor 1.4,number of signals acquired 2,image matrix 312 x 260 and acquisition time 3 min 34 s.SPAIR T2WI:TR/TE 3000/55 ms,slice thickness 3.0 mm,overlap 0.3 mm,FOV 130 × 130 mm,number of slices 45,image matrix 236 x 169 and acquisition time 3 min 36 s.DTI sequence used two b values.Each volunteer underwent three DTI scans.The low b-value of each DTI imaging was 0,and the high b-value was 600、800 and 1000 s/mm2.The remaining parameters of DTI were as follows:TR/TE 14034/94 ms,image matrix 128 x 125,FOV 160 ×60 mm,overlap 0,slice thickness 3.0 mm,flip angle 90°,sensitivity encoding factor 2.5,number of signals acquired 2,number of slices 100,number of directions of motion-probing gradients 15,acquisition plane axial and acquisition time 8 min 15 s.3.Post-processing and quality evaluation of DTI imagesDTI images of bilateral TN and CPN from twelve volunteers with different high b-values were analyzed by two experienced radiologists,blinded to each other.Two observers did not know the b-values of each DTI image.The raw diffusion tensor data were transferred and processed on an independent workstation(Extended MR Workspace,version 2.6.3.2,Philips Healthcare).T1-weighted images were used for image fusion.To visualize the nerves and extract diffusion parameters,tractography was performed on fused images.The image score was based on the display of the nerve fibers and image artifacts.Three points:the nerve fiber bundles were continuous,complete,smooth edges and there were few artifacts in the image.One points:the nerve fiber bundles were discontinuous,the edges were not smooth,and there were heavy artifacts in the image.Two points:image quality between the above two situations.The b-value of a set of high score DTI images is applied to the next step.4.Placement of region of interest(ROI)and measurement of DTI parametersA group of highest score DTI images from last step were post processed on a workstation.Images were analyzed by two radiologists blindly.T1WI and SPAIR T2WI images were used to locate the TN and CPN.T1WI image was fused with DTI image.After the image fusion of T1-weighted image and DTIs,ROIs were placed manually on TN and CPN respectively.The slice through the center of the knee joint space was used for the first ROI placement.The other two ROIs were placed on the upper 10 mm and lower 10 mm slices,respectively.The average measurements of the FA and ADC values in the three ROIs on tensor calculated images were recorded by two radiologists independently,and the average of the two radiologists’ results were used for further analysis.5.Statistical AnalysisThe score of DTI image was expressed as(x±s).Interobserver agreement f’or visualization scores was calculated using the Kappa statistic.Kappa values of 0-0.20.poor agreement;0.21-0.40,midrate agreement;0.61-0.80,good agreement;and 0.81-1.00,excellent agreement.Three DTI image groups with different b-values were evaluated by Kruskal-Wallis and Nemenyi rank sum tests.Statistical analysis was performed using SPSS software(version 17,SPSS Inc.,Chicago,IL).A value of P<0.05 was considered statistically significant.ResultsThe average Kappa values of two observers in evaluation of DTI image quality was 0.793.0.699,and 0.787.And the agreement was good.The average quality scores of three groups of DTI post-processing image with different high b-values were 2.146±0.667,2.688±0.528 and 2.229±0.642.The difference was statistically significant(x 2 =10.672.P=0.005),The score of DTI images with a high b-value of 800 s/mm2 was higher than that of 600 s/mm2(x 2 =8.86,P=0.01)and 1000 s/mm2(x=6.96.P=0.03).There was no significant difference between the score of DTI images with a high b-value of 600 s/mm2 and 1000 s/mm2(x2=0.11,P=0.94).The score of DTI images with a high b-value of 800 s/mm2 was highest(P<0.05).The average FA.ADC values of TN in normal volunteers measured by two observers were 0.593±0.064 and 1.080±0.075×10-3 mm2/s,respectively.The average FA、ADC values of CPN in normal volunteers measured by two observers were 0.623±0.059 and 0.993±0.141×10-3 mm2/s,respectively.Conclusions1.DTI can show the tibial and the common peroneal nerve fiber bundle clearly.2.DTI images with a high b-value of 800 s/mm2 of TN and CPN in healthy adult volunteers had better image quality in our study.Part II Assessment of tibial and common peroneal nerves in diabetic peripheral neuropathy by diffusion tensor imagingObjectiveDiabetic peripheral neuropathy(DPN)is one of the most common complications of diabetes mellitus(DM),and the incidence of DPN varies from approximately 10%to 90%.DPN can involve sensory,motor or autonomic nerves,resulting in loss of limb feeling,pain,and bladder and bowel dysfunction.It begins from the distal end of lower extremity,and gradually developed to the proximal end.Our study was to explore the diagnostic accuracy of DTI parameters in DPN and detect correlations with electrophysiology.Materials and Methods1.Study populationControls:twenty-four lower extremities of 12 healthy adult volunteers(4 men,8 women;mean age 49 years,range 28-65 years)were enrolled from the first part of our study.DPN groups:Twenty lower extremities of 10 patients(3 men and 7 women;mean age 51 years,range 37-65 years)with history of DM and available electrophysiology were enrolled from May to December 2015.Other admission criteria were the same as the first part.2.Electrophysiology examinationBefore the MRI examination,all cases underwent electrophysiology of the lower limbs by experienced examiners.Motor nerve conduction velocity(MCV)and motor nerve conduction amplitude of TN and CPN in each subject were recorded.The normal MCV value is at least 41 m/s in the TN and CPN.The normal amplitude values are at least 4 mV in TN and at least 2 mV in CPN.3.MR examinationAll subjects were examined using a 3.0 T MRI system(Achieva,Philips Healthcare,Best,Netherlands).The subjects were placed supine with feet-first position and unilateral imaging(first right and then left).An eight-channel knee coil was employed at the knee level.MRI sequences included T1WI,SPAIR T2WI and DTI.In sequence of DTI,b-value were 0 and 800s/mm2.Other MR protocols were the same as part I.4.DTI post-processingThe raw diffusion tensor data of patient groups were transferred and processed on an independent workstation.MR images were analyzed by two independent radiologists,blinded to each other.T1-weighted images were fused with DTI images.After the image fusion of T1-weighted image and DTIs,ROIs were placed manually on TN and CPN respectively.The three slices of ROI placement were the same as that of control groups.The average measurements of the FA and ADC values in the three ROIs on tensor calculated images were recorded by two radiologists independently,and the results were used for further analysis.5.Statistical analysisThe numerical variable data was expressed as(x ±s).Electrophysiology results were used as reference standards.The inter-reader agreement of the FA and ADC values was assessed using intraclass correlation coefficient(ICC)analysis.ICC<0.4,poor agreement;0.4<ICC<0.75,common to good agreement;ICC<0.75,excellent agreement.The difference of FA and ADC values in the two groups were evaluated by unpaired t test.Levene test was performed before the t test.Receiver operating characteristic(ROC)analysis was performed to evaluate the diagnostic performance of DTI parameters for TN and CPN.The area under the ROC curve(AUC),cut-off values,sensitivity and specificity were calculated.Pearson correlation coefficient was performed for the analysis of the correlation between DTI parameters and electrophysiology parameters.The strength of the correlation was interpreted as follows:r = 0-0.20,very weak;0.21-0.40,weak;0.41-0.60,moderate;0.61-0.80,strong;0.81-1.00,very strong.Statistical analysis was performed using SPSS software(version 17,SPSS Inc.,Chicago,IL)and MedCalc(version 16.2.4;MedCalc Software,Mariakerke,Belgium).A value of P<0.05 was considered statistically significant.Results ICCs for intra-reader agreement of FA and ADC measurements were 0.895(95%confidence interval(CI),0.807-0.943)and 0.850(95%CI,0.714-0.920)for TN and 0.902(95%CI,0.809-0.948)and 0.914(95%CI,0.842-0.953)for CPN,respectively.The mean FA value of TN in the patient group was 0.534±0.052.The mean FA value of CPN in the patient group was 0.553±0.068.The mean FA value of TN in the control group was 0.593±0.064.The mean FA value of CPN in the control group was 0.623±0.059.The FA values of both TN and CPN in the patient group were significantly lower than those in the control group(F=1.095/0.398,t=3.345/3.660,P<0.05).The mean ADC value of TN in the patient group was 1.173±0.088×10-3 mm2/s.The mean ADC value of CPN in the patient group was 1.128±0.184x 10-3 mm2/s.The mean ADC value of TN in the control group was 1.080±0.075×10-3 mm2/s.The mean ADC value of CPN in the control group was 0.993±0.141×10×10-3 mm2/s.The ADC values of both TN and CPN in the patient group were significantly higher than those in the control group(F=0.118/3.752,t=3.771/2.741,P<0.05).The cut-off value obtained by ROC analysis was 0.58 for FA values(sensitivity,80%;specificity,58.3%;AUC,0.755)and 1.15 × 10-3 mm2/S for ADC values measurement(sensitivity,70%;specificity,83.3%;AUC,0.806)of TN.For CPN,the cut-off value was 0.59 for FA values(sensitivity,80%;specificity,70.8%;AUC,0.771)and1.15×10 3 mm2/s for ADC values measurement(sensitivity,50%;specificity,87.5%;AUC,0.711).In the patient group,there was moderate correlation between DTI parameters and MCV.The FA values positively correlated with MCV(r=0.487/0.460,P<0.05),and ADC values negatively correlated with MCV(r=-0.592/-0.479,P<0.05).There were no significant correlations between amplitude and DTI parameters in the patient group(P>0.05).Conclusion1.FA values of TN and CPN in control group are significantly lower than those in DPN group.ADC values of TN and CPN in control group are significantly higher than those in DPN group.2.There is excellent interobserver performance in DTI measurements of TN and CPN in healthy adult volunteers.3.DTI parameters demonstrate moderate diagnostic accuracy in detection of DPN.4.There is moderate correlation between DTI parameters and MCV in DPN patients. |
PDF Full Text Request