【作者】 邢古生；

【导师】 周纯武；

【作者基本信息】 中国协和医科大学 ， 影像医学与核医学， 2009， 博士

【摘要】 目的:本研究通过非酒精性脂肪肝背景下的肝细胞肝癌(HCC)的CT及MR扫描,对化学位移T1WI双回波成像进行定量评估,并与CT所见对比;并总结在脂肪肝背景下肝细胞肝癌的T1WI表现特征。材料与方法:2008年6月至2009年2月,对临床疑似或HCC高危人群行多排螺旋CT及3-T MR扫描,排除脂肪肝以外的肝脏背景异常后,共46例HCC患者纳入本研究,其中男性39例,女性7例,年龄35-76岁(平均50.53岁)。本研究患者被分为两组:非酒精性脂肪肝组(11例)和肝脏背景无异常组(35例)。所有患者有完整的多排螺旋平扫图像、MR化学位移T1WI正反相位图像,以及完整的临床资料。对CT及MR诊断脂肪肝进行一致性检验,并采用AFROC曲线分析,对比CT及T1WI双回波图像定量评估脂肪肝背景的效能,并对比不同肝脏背景下HCC病灶的对比噪声比(CNR)变化特征。结果:1、所有46例患者中,在评估脂肪肝程度方面,T1WI双回波肝脏信号强度指数的大小与CT扫描的肝、脾CT值之比呈显著负相关关系,相关系数r=-0.487,P=0.007;T1WI双回波肝脏信号强度指数与肝脾的CT值之差也呈显著负相关关系,相关系数r=-0.474,P-0.007。T1WI双回波与CT值在反映脂肪肝的程度方面,其评估具有一致性。2、T1WI双回波肝脏信号强度指数诊断脂肪肝的ROC曲线下面积(Az)为0.950±0.031。CT扫描肝脾CT值比值诊断脂肪肝的Az为0.938±0.056,肝脾CT差值的Az为0.940±0.053;三者相比差异无显著性(P值分别为0.85、0.87、0.44)。3、在T1WI正相位图像,无脂肪肝背景组HCC病灶的CNR绝对值略低于脂肪肝组(-47±8 VS-48±14.1),差别无显著性(P=0.946);反相位两组病灶CNR值分别为-28.12±7.04、-9.67±11.41,虽然CNR绝对值存在差异,但差别仍无显著性意义(P=0.175)。4、T1WI双回波图像中,正常肝脏背景组与脂肪肝组HCC病灶的CNR变化差值分别为-18.22±3.98、-38.40±8.86,脂肪肝组病灶CNR变化较对照组显著(P=0.023)。结论:在HCC疑似或高危人群中,MR化学位移T1WI双回波图像对脂肪肝的评估与CT一致,均有很高的诊断效能;HCC伴有脂肪肝时,病灶在T1WI双回波图像的显示会发生显著变化。第二部分CT三期增强与MRI多期动态增强扫描诊断肝细胞肝癌的对比分析目的:比较多排螺旋CT三期增强、3-T MR多期动态增强扫描对HCC的诊断效能;对比CT单动脉期与MR双动脉期扫描对HCC病灶的显示能力。材料与方法:2008年6月到2009年2月,共60例临床疑似或HCC高危患者纳入本研究,并完成CT及MR检查。39例确诊为HCC,共59个HCC灶,其中男性35例,女性4例,年龄35—74岁(平均56.1岁)。CT采用平扫加三期增强,MR采用LAVA技术双动脉期、双门脉期、双静脉期、2—3分钟延迟期,共7期动态增强扫描。通过对三位观测者的CT和MR观测数据行Alternative-free-responseROC(AFROC)曲线分析,并计算曲线下面积(Az),比较两种方法的诊断效能。另外单独比较CT动脉期和MR双动脉期对HCC病灶的显示能力。结果:1、CT增强扫描的Az平均值0.812±0.0118,MR多期动态增强扫描的Az平均值0.9093±0.0072,差别均无显著性意义(P值分别为0.15、0.14、0.09)。在＜1cm的HCC病灶组,CT和MR增强扫描的平均敏感性为63.9%和80.5%,MR高于CT扫描(P=0.013)。在1—3cm组和＞3cm组,CT和MR增强扫描诊断的敏感性差别均无显著性(P＞0.05)。在所有病灶中,CT多期增强扫描的敏感性低于MR动态增强扫描(83.6%VS 88.7%),差异无显著性(P＞0.05),CT阳性预测值(PPV)也低于MR扫描(93.1%VS 96.3%),差异无显著性(P＞0.05)。2、MR动脉早期对病灶检出敏感性和PPV分别为67.8%和93.6%,MR动脉中期与动脉双期扫描的敏感性和PPV均相同,分别为84.6%和87.9%;MR动脉早期对HCC病灶的总体检出敏感性低于动脉中期和双动脉期(P=0.034)。CT动脉期对病灶检出的敏感性和PPV分别为76.6%和92.9%,CT动脉期对病灶的检出敏感性低于MR双动脉期,但差异无显著性意义(P=0.073);CT动脉期的阳性预测值高于MR双期,差异无显著性意义(P=0.58)。3、CT动脉期和MR双动脉期对＜1cm组HCC病灶的敏感性平均值分别为55.5%、75%,MR双动脉期的敏感性显著高于CT动脉期(P=0.036);在1—3cm组,CT动脉期与MR双动脉期的敏感性分别为71.2%、82.7%,但差异无显著性意义(P=0.24);在＞3cm组,CT动脉期与MR双动脉期的敏感性分别为95%、96.7%,差异无显著性意义(P=0.21)。结论:对小于1cm的病灶,MR增强扫描的敏感性高于CT扫描;大于1cm的病灶,两者的诊断敏感性无显著差别。病灶体积越大,MR双动脉期、CT动脉期显示病灶的敏感性越高,但其敏感性和阳性预测值无显著差异。MR双动脉期扫描时,动脉早期扫描诊断价值有限。第三部分CT及MR多期动态增强扫描对肝细胞肝癌介入后随诊的对比分析目的:比较多排螺旋CT与MR多期动态增强扫描对肝细胞肝癌(HCC)经肝动脉化疗栓塞(TACE)后肿瘤局灶复发或残存的诊断价值。材料与方法:18例结节型HCC患者共38个病灶,TACE介入后45天到11个月分别行多排螺旋CT增强扫描及MR多期增强扫描,检查平均间隔9天(1-21天)。两位影像诊断医生分别读片,评价肿瘤复发或残存。所有病灶经临床化验、血管造影、TACE后CT随诊等方法证实。通过计算两种诊断方法的ROC曲线下面积(Az),比较两者诊断的准确性。并评价CT及MR诊断的敏感性差异。结果:所有38个病灶中,24个(63.2%)有肿瘤残存或复发。两位观测者的诊断评分具有较高的一致性(κ值为0.732,P＜0.05)。CT动态增强ROC曲线下面积平均为0.751,MR多期增强ROC曲线下面积平均为0.949,差异具有显著性(两位观测者的P值分别为0.037、0.050),表明MR多期增强扫描诊断的效能高于CT增强扫描。CT诊断HCC病灶复发或残存的敏感性和阳性预测值(PPV)平均分别为43.8%和100%;MR诊断的敏感性和PPV平均值分别为84.5%和93.3%;MR诊断敏感性显著高于CT扫描(两位观测者P值分别为0.006和0.005)。结论:在评估肝细胞肝癌TACE治疗后肿瘤残存或复发方面,MR多期增强扫描优于CT增强扫描。更多还原

【Abstract】 Part IHepatocelluar Carinoma with Nonalcoholic Fatty Liver Background:Quantitative Evaluation of MR Double-echo Chemical Shift T1 Weighted Images and Compared with CT ImagesPurpose:CT and MR scan were performed in patients with hepatocellular carcinoma, the fatty liver background and HCCs were quantitatively evaluated by chemical shift double-echo T1 weighted imaging,and compared with CT scan.Materials and Methods:Between June 2008 and February 2009,the suspected and risk patients with HCC were performed multiple-detector CT and 3-T MR scan.The patients with liver background abnormal were ruled out except for nonalcoholic fatty liver.A total 46 HCC patients were fit into our study(male 39,female 7,ranged 35-76 years old,mean age 50.53 years).The HCC cases were divided into two groups,the fatty liver(11 cases) and normal background of liver(35 cases).The images and clinical data were available in all patients.The ROC curve was used to compared the ability of quantitative evaluation for fat content in the liver at CT and dual-echo T1WI images.The different HCCs features of contrast noise rate(CNR) in fatty liver were also compared at MR images.Result:1.To evaluate the fat content in the liver,the signal intensity(SI) index of dual-echo T1WI showed a linear negative correlation with liver-to-spleen CT value ratio,the correlation was significant(r=-0.487,P=0.007).A significant linear negative correlation (r=-0.474,P=0.007) was also found between the SI index of dual-echo T1WI and liver-to-spleen CT value difference.2.For the evaluation of fatty liver,the area under ROC(area index Az) value of dual-echo T1WI was 0.950±0.031,the Az value of liver-to-spleen CT ratio was 0.938± 0.056,the Az value of liver-to spleen CT value difference was 0.940±0.053.Compared between any two means,the differences were not significant(the P values were 0.85, 0.87,0.44,respectively).3.Compared with the fatty liver group,the CNR of HCCs in control group was higher(-47±8 VS -48±14.1) at the in-phase MR images,but the difference was not significant(P=0.946).At the opposed images,the CNR values of HCCs were -28.12±7.04 and -9.67±11.41,respectively,the difference was not significant(P=0.175) too.4.At the dual-echo T1WI,the mean CNR variation of HCCs in control group and study group was -18.22±3.98 and -38.40±8.86,respectively.The difference was significant(P=0.023).Conclusion:The quantitative evaluation of fat content in the liver for the patients with HCC,the dual-echo T1WI appeared a consistent with CT.The intensity of HCCs at dual-echo T1WI appeared a significant change in the fatty liver background. Purpose:Triple-phase enhancement CT and multiple-phase dynamic-enhancement MR scan were performed in patients with hepatocellular carcinoma(HCC).The diagnostic ability of the two methods was compared.The detectability of dual-arterial phase of MRI and single-artery phase of CT scan for HCCs was evaluated.Materials and methods:Between June 2008 and February 2009,a total of 60 patients suspected or risk of HCC were performed CT and MR scan.59 HCCs were confirmed in 39 patients(male 35,female 4,ranged 35-74 years old,mean aged 56.1 years).CT was performed with routine CT scan and a triple-phase enhancement scan. MR imaging was performed with LAVA technique,the images included mask images, dual-artery phases,dual-portal phases,dual-venous phases and 2-3 minute delayed phase. Three observers evaluated the CT and MR imaging separately,and the results were compared with alternative-free-response ROC(AFROC) curve,the area under ROC(Az) was calculated to compare the diagnostic ability.In addition,the detectability of dual-artery phase of MR images was compared with the single-artery phase of CT images.Result:1.The Az value of CT for the diagnosis of HCC was 0.812±0.0118,the Az value of MR was 0.909±0.0072,the difference was not significant(P value of three observers was 0.15,0.14,0.09,respectively).In the group of less than 1cm in the diameter of HCCs,the sensitivity of CT and MR was 63.9%and 80.5%,the difference was not significant (P=0.013).In the groups of 1-3cm and larger than 3cm of HCCs,the sensitivity of CT and MR appeared no significant difference(P＞0.05).Of all HCCs,the sensitivity of CT was lower than MR(83.6%VS 88.7%),but the difference was not significant(P＞0.05). The positive predict value(PPV) of CT was also lower than MR(93.1 VS 96.3%),but the difference was not significant too(P＞0.05).2.The mid-artery phase had the same sensitivity and PPV with dual-artery phase of MR of 84.6%and 87.95,respectively.The early-artery phase of MR appeared a significantly lower sensitivity of 67.8%(P=0.034),PPV of early-artery phase imaging was 93.6%.The artery phase of CT scan had a sensitivity and PPV of 76.6%and 92.9%, respectively.The sensitivity was lower than double-artery phase of MR imaging,and the PPV was higher,but the difference was not significant(p=0.58).3.In the group of less than 1cm in diameter of HCCs,the sensitivity of CT single-artery phase images and MR dual-artery phase images was 55.5%and 75%, respectively,the later showed a higher sensitivity(P=0.036).In the groups of 1-3cm and＞3cm,the sensitivity of the two technique compared no significant difference(71.2% VS 82.7%,P=0.24.And 95%VS 96.7%,P=0.21).Conclusion:The diagnostic ability of triple-phase enhancement CT and multiple-phase dynamic-enhancement MR scan for HCCs was similar.For the lesions less than 1cm in diameter,the dynamic-enhancement MR was superior to that of contrast-enhancement CT scan.While for the larger ones,the difference was not significant.For the detection of HCCs,the early-artery phase MR images did not provide additional benefit for dual-artery phase enhanced MRI detectability. Purpose:To compare the detectability of MR Dynamic enhancement with that of Contrast-enhanced CT for the detection of local hepatocellular carcinoma(HCC) recurrence post-transcatheter arterial chemoembolization(TACE).Materials and methods:Eighteen patients with 38 HCCs underwent MR and spiral CT scan with a mean interval of 9 days(range,1-21days) after 45 days to 11 months of TACE.The enhanced MRI examination was performed with a LAVA technology on a 3-T unit.Both MRI and CT scan included multiphase imaging.Two observers independently interpreted the CT and MR images blindly.The diagnostic ability of the two techniques was evaluated by comparing the area under the ROC curve(Az value).The sensitivity was also evaluated.Results:Of all 38 HCCs,24(63.2%) lesions were confirmed local recurrence after TACE.The observers appeared a high consistent with the diagnosis(a kappa value of 0.732,P＜0.05).The Az value of contrast-enhancement MRI was higher than that of dynamic enhanced CT scan(0.949 VS 0.751,a P value was 0.037 and 0.050,respectively. The diagnostic ability of MR LAVA was higher than that of dynamic CT scan for the local HCCs recurrence.The sensitivity of contrast-enhancement MRI for the detection of recurrence was higher than CT scan(84.5%vs 43.8%,P＜0.05).Conclusion:Contrast-enhancement MRI has a high diagnostic ability of detecting local recurrence of HCC post transcatheter arterial chemoenbolization comparing contrast-enhanced CT scan.更多还原