Clinical Study Of 3T MR Functional Imaging In Evaluating Liver Fibrosis

BackgroundLiver fibrosis is the wound-healing response to various liver damage,in the form of a progressive accumulation of fibrosis.It alters the tissue structure and function of liver,leading to cirrhosis and liver failure.Many causes can induce hepatic fibrosis.In our country,however,chronic viral infection is the main cause.It is reported in some researches that liver fibrosis can be reversed through effective treatment;but if the cause stays on,it would develop into liver cirrhosis.So,early detection of liver fibrosis and cirrhosis has important clinical implications for the determination of antiviral treatment options and patient prognosis.Chronic liver diseases progress through different histological stages to final cirrhosis,and histological stages serve as a key predictor in the clinical management of patients with chronic hepatitis.Thus,the diagnosis of the stage of liver fibrosis and the grade of inflammation in patients with chronic hepatitis is essential for therapeutic and prognostic evaluation.The gold standard in assessing fibrosis is still represented by liver biopsy, despite the invasiveness of the procedure and the likelihood of sampling errors. Therefore it is an urgent need to search for a noninvasive diagnosis method to quantify liver fibrosis.MRI has become an increasingly important imaging modality for the investigation of patients with chronic liver disease.Some functional imaging methods have been investigated for this purpose,but their clinical role remains undefined.To our knowledge,however,the comparison among these methods in the diagnosis of liver fibrosis has not been reported.Thus,the purpose of our study is to prospectively evaluate the sensitivity and specificity of various functional imaging methods by using 3.0 T MR in the diagnosis of liver fibrosis(such as DWI,Dynamic contrast-enhanced MRI,MTI and T2mapping),in order to choose reasonably in clinical work.Part One:Clinical Study of Diffusion-weighted MR Imaging in Evaluating Liver FibrosisObjectiveThe purpose of this study is to optimize the technical parameters of diffusion-weighted MRI in liver with 3T MR scanner and to evaluate its ability for evaluating of liver fibrosis quantificativly.Materials and Methods1.SubjectsA total of 64 cases were incorporated into this study and were divided into 6 groups according to the stage of liver fibrosis:S0(14 cases),S1(9 cases),S2(14 cases),S3(7 cases),S4(3 cases),liver cirrhosis(17 cases).A total of 46 cases were divided into 5 groups according to the grade of inflammation:G0(14 cases),G1(2 cases),G2(11 cases),G3(16 cases),G4(4 cases).2.Instruments and equipmentDWI of the liver was performed on a GE Signa Excite 3.0-T superconductive MRI scanners with an eight-element phased-array superficial coil.3.Examination methods and parametersThe patients and control subjects were asked to fast for 4 hours before the study. Breathhold diffusion-weighted imaging of the liver was performed,using single shot echo-planar imaging sequences in conjunction with parallel imaging with an acceleration factor of 2.0Ph.ASSET correction scan was performed before DWI imaging with scan range from the middle chest to the lower abdomen.Four breath-hold acquisitions were obtained in the same liver location at b values of 0-300,0-600,0-800 and 0-1000s/mm².The following parameters were used: TR/TE＝1500/46.7～59.2ms,NEX＝4,slice thickness＝8mm;gap＝2mm;matrix＝128×128,field of view＝38×38cm.First of all,18 healthy volunteers were imaged twice under each b value with diffusion gradient applied along the section-select direction(S/I) and in three orthogonal directions(All).The rest of examinees were imaged with"ALL"model with other imaging parameters invariably.4.Data processing and data analysisDiffusion analysis software package on a workstation(GE ADW4.3) was used to obtain ADC maps for each b value.The quality of DWI images obtained from two patterns of diffusion gradient directions was evaluated by 5 indicatrixs as follows: Whether the intrahepatic pipe has a clear border;Whether the magnetic artifacts were sensitive,Whether the signal of liver parenchyma is homogeneous and fine,whether the left lobe of liver shows better,and how much the background noise is.Three scores(2,1,0) were given,representing respectively three levels as excellent,good, bad.Signal intensity(SI) was measured on the images recorded in a homogeneous circular(area,1.3cm2) at the right posterior lobe of the liver and the noise of background.Signal to noise ratios(SNR) of liver on DWI were calculated:SNR＝SI/SDnoise,in which SDnoise is the standard deviation of the background noise. During the process,great care needed to be taken to ensure the consistency of ROI placement at different diffusion gradient directions.ROIs were placed in areas that were as similar as possible to the areas assessed in two patterns.Quantitative ADC maps were derived automatically on a voxel-by-voxel basis by using software.Three regions of interest(ROIs) were placed in the posterior right hepatic lobe per slice(mean size＝1.3 cm) on three consecutive slices to measure liver ADC values.Care was taken to exclude vessels and bile duct.The final ADC per subject used for analysis was the average of nine measurements.With method mentioned above,the ADC values were measured twice by examining the b value of the DWI images of 13 examinees.Differences were then compared. 5.Statistical analysisSPSS 11.5 statistical software was used for analysis.All values were expressed as mean±standard deviation.For all tests used,A p value＜0.05 was considered statistically significant.The paired Kruskal-Wallis test was used to compare the images' quality score obtained with two Pattern that diffusion gradient directions were applied.The paired T test was used to compare the difference of SNR and the ADC value of the two DWI images with two patterns.The Spearman correlation analysis was used to study the relevance between the b value and signal strength,signal to noise ratio and ADC value on DWI images.The paired T test compared the ADC values measured two times from the same patients.The Spearman rank correlation test was used to assess the correlation between liver ADC and stage of fibrosis and grade of inflammation.One-way ANOVA test was used to compare the difference of ADC value among fibrosis stage and inflammation grade under the conditions of different b values,and LSD method was used to perform multiple comparisons.Receiver operating characteristic analysis was used to assess the performance of ADC in prediction of the presence of stage 1 or greater,stage 2 or greater,stage 3 or greater of liver fibrosis and grade 2 or greater, grade 3 or greater of inflammatory activity.Result1.DWI MethodologyDWI image quality obtained with diffusion gradient applied in three orthogonal direction was significantly better than that obtained with diffusion gradient applied along the section-select direction(P＜0.05) at each b values.The SNR of the image with"all"pattern tended to be higher than those with"S/I"pattern and the difference was significant(P＜0.05) at b＝300,800 s/mm2.ADC value tended to be higher than the S/I image and the difference was significant at b＝600,800 s/mm2(P＜0.05).There was negative correlation between b value and signal strength,SNR and ADC values of DWI images(P＜0.01).There were no significant differences between ADC values measured two times based on the ROI selection criteria(P＞0.05).2.ADC Values and Fibrosis StageThere was moderate but significant negative correlation between ADC and fibrosis stage.The best correlation was observed for the ADC at b values 600 and 800 s/mm²(r＞0.5).There was a decreasing trend in hepatic ADC with increasing degree of fibrosis. There were significant differences among different fibrosis stages at each b values(P＜0.01).In multiple comparisons,there were no significant difference between S0 and S1 for all b values and there were significant differences between stage 0 and stage 2-4(P＜0.05).In addition,there were significant differences between stage 1 and stage 3 fibrosis,stage 1 and cirrhosis for b value of 800～1000 s/mm²,between stage 1 and stage 2 and cirrhosis for a b value of 600～800 s/mm².Using ROC analysis,we found hepatic ADC to be a significant predictor of stage 1 or greater fibrosis with an AUC of 0.733,0.844,0.793,0.694 at b value of 300,600,800,1000s/mm².AUC were 0.748,0.835,0.820,0.775 when hepatic ADC was used to predict S≥2 and were 0.748,0.855,0.840,0.791 respectively when hepatic ADC was used to predict S≥3.The largest AUC was observed at b value of 600s/mm² and followed at b value of 800s/mm².At b value of 600s/mm²,the diagnosis of S≥1 accuracy of ADC values was 84.4%.When assuming ADC values＜1.453×10^-3mm²/s as the standard,the sensitivity was 75%and specificity 73.9%;the diagnosis of S≥2 accuracy was 83.5%,When assuming ADC＜1.453×10^-3mm²/s for the diagnosis of community value,the sensitivity was 80.5%and specificity 73.9%;predicted moderate and above fibrosis (S3 or grater) for the area under the curve was 0.855,when assuming ADC value≤1.414×10^-3mm²/s as the standard,sensitivity was 81.5%,specificity was 73%.3.ADC Values and Inflammation GradeThere was a decreasing trend in hepatic ADC with increasing degree of inflammation.The correlation between ADC and inflammation grade(P＜0.05) varied from weak to moderate.There were a significant difference at b values of 600,800 s/mm² among different degree inflammation(P＜0.05).Using ROC analysis,we found ADC to be a significant predictor of grade 2 or greater inflammation with an AUC of 0.783,0.832 respectively(b＝600,800 s/mm²) and to be a significant predictor of grade 3 or greater inflammation with an AUC of 0.786,0.782 respectively.For prediction of great 2 or greater inflammation at b values of 600 s/mm²,we found an AUC of 0.783 with a sensitivity of 75%,specificity of 73.9%for a hepatic ADC of 1.453×10^-3mm²/s or less.For prediction of great 3 or greater inflammation at b values of 800 s/mm²,we found an AUC of 0.783 with a sensitivity of 75%,specificity of 73.9%for a hepatic ADC of 1.433×10^-3mm²/s or less.Conclusions1.When conducting liver DWI,the selection of diffusion-sensitive gradient direction has a certain impact on image quality and ADC value measured.In this study,diffusion-sensitive gradient were applied in three orthogonal directions,a better image quality and stability of the ADC value were obtained.2.There was negative correlation between ADC and fibrosis stage,inflammation grade.ADC values have a certain contribution to the evaluation of the stage of hepatic fibrosis and inflammatory activity.The former presents a higher accuracy.3.The diagnosis performance of ADC values in predicting the stage of liver fibrosis was at the b of 600 s/mm².4.DWI imaging is a non-invasive technology and could be operated easily,and ADC value measurement has better reproducibility.DWI could provide a reference in early diagnosis,treatment and follow-ups for chronic patients.Part Two Application Study of Dynamic Contrast-Enhanced MRI of Whole Liver in Evaluating Liver FibrosisObjectiveTo analyze the homodynamic changes of liver fibrosis by using enhanced dynamic MRI;to investigate the role of Dynamic Contrast-Enhanced MRI in evaluating liver fibrosis. Materials and Methods1.SubjectsThirty-one consecutive patients were prospectively enrolled in this study.They are divided into 4 groups according to the stages of fibrosis:normal group(SO 9 cases),mild fibrosis group(S1 1 case,S2 2 cases),moderate to severe fibrosis group (S3 1 case,S4 3 cases),15 cases of liver cirrhosis group.2.Equipment and contrast mediumMagnetic resonance scanner and the surface coil were the same as used in Part One.Gd-DTPA(Dimeglnmine Gadopentetate Injection,Kangchen,China) was injected as paramagnetic contrast medium with high pressure injection(Ulrich corp).3.Examination TechniqueWhole-liver enhanced dynamic MR imaging was performed by using a liver acceleration volume acqui-sition technique in Axial plane.Phase-accelerating factor was 2.5Ph.The following imaging parameters were used:2.8/1.2(repetition time msec/echo time msec),5.4-cm slice thickness resulting in an interpolated 2.7-mm section thickness,and 125 Hz bandwidth.The starting time of acquisition of dynamic contrast-enhanced images was synchronized with the start ofâ…£bolus administration of 0.1 mmol/kg of gadopentetate dimeglumine followed by a 20ml saline flush injected at a rate of 4 ml/sec with an MR-compatible power injector.A total of four scanning were conducted,each including three phases,and each phase lasted 8s,the total scan time totaled approximately 160s.Patients were requested to hold their breath for 24s thrice.The first breath-holding was started simultaneously with administration of gadopentetate dimeglumine and the interval between breath-holding periods was 10,20,35sec,respectively.4.Image AnalysisThe images were transferred to GE ADW4.3 workstation and were analyzed using the SER software of Functool software package.Regions of interest(ROIs) were drawn manually on the main portal vein,the proximal abdominal aorta,right lobe areas of the liver and spleen.Time-signal intensity curve(TIC) generated automatically.Signal intensity curve peak(Sip),the baseline signal intensity value (SI₀) and peak time(time to peak,TTP) were recorded and peak height of the TIC, the largest increase in the slope of the signal(maximum slope of increase,MSI) and signal the largest drop in the slope(maximum slope of decrease,MSD) were calculated.MSI was calculated as follows:MSI/MSD＝(SI₂-SI₁) / t,where"SI₂"and"SI₁"is the signal intensity at the point of biggest signal intensity changes in two adjacent phase on an increase or decrease section of TIC,and"t"is the interval time of two adjacent phase.The 1 to 3 phases were defined as the hepatic artery phase,and 4～6 phase were defined as the portal vein-phase.5.Statistical AnalysisStatistical software(SPSS,version 11.5) was used for all statistical computations. For all tests used,A p value＜0.05 was considered statistically significant.The Spearman rank correlation test was used to assess the correlation between TIC parameters and stage of fibrosis and grade of inflammation.One-way ANOVA test was used to compare the difference of TIC parameters among fibrosis stage and inflammation grade and LSD method was used to perform multiple comparisons. Receiver operating characteristic(ROC) curve analyses were conducted to evaluate the utility of the TIC parameters for the prediction of fibrosis stage≥1,≥2,and≥3, and for the prediction of inflammation grade≥2.Results1.There was moderate but significant negative correlation between the peak height of portal vein,liver,spleen and fibrosis stage(P＜0.01).There was significant positive correlation between the peak time and fibrosis stage(P＜0.01).There was significant negative correlation between portal MSI,hepatic arterial phase,MSI, hepatic portal venous phase MSI,liver MSD,spleen MSI,spleen MSD and fibrosis stage(P＜0.01).However,There was no significant correlation between TIC parameters and the degree of stage of inflammation,only spleen peak height has a negative correlation with it(P＜0.01).2.There was a significant difference among patients who have different degrees of liver fibrosis((P＜0.05)) in peak height,TTP and MSI of portal vein.There was a significant difference among patients who have different degrees of liver fibrosis in TIC Parameters of liver and spleen(P＜0.05).3.Receiver operating characteristic analysis showed that the relative good estimated parameters used to predict S≥1 were liver MSI of Portal venous phase,spleen MSI of Arterial phase,spleen MSD(area under the receiver operating characteristic curve,0.747,0.738,0.783,respectively).For prediction of stage lor greater fibrosis with MSD of spleen,we found an AUC of 0.783 with a sensitivity of 81.8%,specificity of 77.8%(MSD≤11.475).All TIC parameters were significante when to diagnosis S≥3(P＜0.05).The estimated paramerers were as follows:liver MSI of arterial phase,the portal vein MSI,liver MSI of portal venous phase,liver MSD,splenic MSI of arterial phase,splenic MSD with area under the curve 0.728～0.877.For prediction of stage 3or greater fibrosis with MSD of spleen,we found an AUC of 0.877 with a sensitivity of 96.7%,specificity of 83.3%(MSD≤11.475).Conclusion1.There was moderate correlation between TIC parameters and stage of liver fibrosis.TIC paramerers can be used to predict advanced liver fibrosis stage.2.The significant indicators of TIC was spleen MSD,splenic MSI,liver MSD, MSI of hepatic portal venous phase and portal vein MSI followed by AUC descending.It is necessary to observe the hemodynamics of the spleen.3.Multiphase dynamic contrast-enhanced MRI with coverage of the entire liver could be used to evaluate the hemodynamic changes in liver,while satisfying the need of morphological observation,and it would be regarded as a practical non-invasive functional imaging method in evaluating hepatic fibrosis.Part Three Magnetization transfer,T2 Mapping Technology in the Diagnosis of Liver FibrosisObjectiveTo evaluate the relevance between MTR values,T2 values and stage of liver fibrosis,grade of inflammatory and to explore the value of MTI and T2mapping technology to predict liver fibrosis.Materials and Methods1.SubjectsA total of 65 cases in which the subjects had done MTI and were incorporated into the analysis eventually.They were divided into 6 groups according to the degree of fibrosis:S0(n＝13),S1(n＝8),S2(n＝15),S3(n＝7),S4(n＝3),liver cirrhosis (n＝19);and were divided into 4 groups by degree of inflammation:normal group (G0 period of 13 cases),mild inflammation group(1 case G1,G2 of 11),moderate inflammation group(G3 of 16 cases),severe inflammation(G4 of 4 cases).A total of 55 cases in which the subjects had done T2mapping imaging were incorporated into the analysis eventually and divided into 6 groups according to the stage of fibrosis:S0(n＝10),S1(n＝7),S2(n＝16),S3(n＝5),S4(n＝3),cirrhosis (n＝14).They were divided into 4 groups according to the grade of inflammation: normal group(G0 period of 10 cases),mild inflammation group(1 case G1,G2 of 11), moderate inflammation group(G3 period of 16 cases),severe inflammation(G4 of 3 cases).2.Apparatus and EquipmentThe same with the Part One.3.Methods and ParametersThe patients and controlled subjects were asked to fast for 4 hours before the study.MTI was performed using SPGR sequence shaped MT saturation pulse.The parameters of SPGR were as follows:TR/TE:76ms/4ms,flip angle:15 degrees,slice thickness:6mm and layer spacing 2mm,matrix 192×192,FOV38×38cm,the receiver bandwidth was 31.3KHz.MT pulse was performed with offset frequency 1200Hz,the effective flip angle 670°and pulse duration 8s.The scanning of the bottom two layers of hilar level position is conducted with the subjects holding breath after expiration.The whole scanning process takes 17s.Inversion recovery fast spin-echo techniques and respiratory triggering technique were used to conduct T2mapping image.T2mapping were acquired using a preparatory time of 35.4 msec followed by an echo train length(ETL) of 8(22.6,33.9,45.2,56.6,67.9,79.2,90.5,101.8ms).FOV:40×40cm,receiver bandwidth: 15.63KHz,slice thickness of 8mm,layer spacing of 2mm,20 slices were acquired using one acquire.Scan time was 7 minutes and 8 seconds.4.Post-Processing & Analysis of ImagesMTI imaging data were analyzed by the magnetization transfer analysis software package.ROI was placed in the right posterior lobe of liver with large blood vessels and edge of organs excluded.MTR value was calculated as(pulse sequence imposed -is not imposed pulse sequences)/pulse sequence imposed.MTR value was obtained automatically by using the software.T2 color map would be obtained by using T2map analysis software within the Functool package tool package.Each pixel value level in the color map represented T2 relaxation values(RV) respectively.The ROIs(area＝1.8mm2) were chosen in the right posterior lobe of the liver of continual 5layers and the average of all ROIs was calculated.5.Statistical AnalysisStatistical software(SPSS,version 11.5) was used for all statistical computations. For all tests used,A p value＜0.05 was considered statistically significant.Spearman correlation analysis was used to study the correlation between liver magnetization transfer ratio(MTR),T2 values and the stage of liver fibrosis,grade of inflammation.Receiver operating characteristic(ROC) curve analyses were conducted to evaluate the utility of the MTR value for the prediction of fibrosis stage≥1,≥2,and≥3.Results1.There was mild but significant positive correlation between MTR value and the stage of liver fibrosis,the grade of inflammation(P＜0.05,r＝0.343,0.347). Receiver operating characteristic analysis showed that the MTR value could be used to diagnosis liver fibrosis but had poor accuracy.The AUC were 0.692,0.683,0.667 respectively when MTR value was used to predict S≥1,S≥2 and S≥3.However, MTR value was not a statistical significance parameter in predicting the grade of inflammation(P＞0.05). 2.There was no significant correlation between T2 relaxation value and the stage of liver fibrosis,the grade of inflammation(P＞0.05)Conclusion1.There is a positive correlation between MTR value and the stage of hepatic fibrosis and grade of inflammation.MTR values can be used to predict the stage of hepatic fibrosis,but the diagnostic accuracy is not high.2.T2mapping with respiratory triggering technique can be used to obtain the T2 relaxation value of the entire liver.But There was no significant correlation between T2 relaxation value and the stage of liver fibrosis,the grade of inflammation.Partâ…£Comparison of Diagnosis Effects on Liver Fibrosis Between DWI and Enhanced Dynamic MRIObjectiveTo compare the effectiveness of DWI and enhanced dynamic MRI in evaluating liver fibrosis.Materials and MethodsA total of 23 patients(S0 8 cases,S1 1 cases,S2 3 cases,S3 1 cases,S4 2 cases, liver cirrhosis 8 cases) who underwent DWI and enhanced dynamic MRI imaging at the same time.ROC curve was used to compare the effect of TIC parameters and ADC value in diagnosing the stage of liver fibrosis.ResultsUsing ROC analysis,we found the TIC parameter was not a significant predictor of stage 1 or greater fibrosis,however,hepatic ADC worked as a good predictor with larger AUC(0.925).TIC parameter and ADC values were both significant predictors of stage 3 or greater fibrosis,however,ADC values were still having largest area under the curve of more than 0.9.The AUC of TIC parameters were 0.765～0.871.ConclusionDWI method is simple without using contrast agent and its diagnosis effect is higher than TIC parameters of enhanced dynamic MRI -- especially in the diagnosis of mild liver fibrosis.However,each method has its own privilege,so the application should be based on the clinical work and the actual need of patients.