BACKGROUND The treatment technology of liver cancer is progressing.In addition to traditional surgical resection,combined therapies of immunotherapy based on immune checkpoint inhibitors,chemotherapy,and transcatheter...BACKGROUND The treatment technology of liver cancer is progressing.In addition to traditional surgical resection,combined therapies of immunotherapy based on immune checkpoint inhibitors,chemotherapy,and transcatheter arterial chemoembolization for hepatocellular carcinoma are more and more widely used.Accurate preoperative diagnosis of liver cancer will provide important information for comprehensive treatment and prognosis evaluation of liver cancer.Sonazoidcontrast-enhanced ultrasound is not only helpful for the qualitative diagnosis of liver lesions,but also has great potential in the diagnosis of histological differentiation of liver cancer.AIM To assess the differentiation of hepatocellular carcinoma(HCC)by utilizing the parameters and imaging features of Sonazoid-contrast-enhanced ultrasound(CEUS).METHODS A retrospective analysis was conducted on the CEUS data of 239 lesions through case-control study.These patients received Sonazoid-CEUS within one week before surgery and were confirmed as HCC by postoperative pathology.Within the cases,patients were further categorized into well-differentiated and poorlydifferentiated group.Time-intensity curves of the region of interest in both arterial and Kupffer phases were generated,allowing for the acquisition of quantitative parameters to assess the diagnostic efficacy in distinguishing lesions between these two groups and determining an appropriate cut-off value.RESULTS Univariate analysis showed that the absolute value of enhancement intensity(EIAV),intensity ratio(IR)and intensity difference(ID)in Kupffer phase were statistically different between the groups with different degree(P=0.015,P=0.000,P=0.000).The sensitivity and specificity were 40.2%,82.4%,80.4% and 78.1%,86.9% and 74.5%,respectively,for differentiating HCC lesions with EIAV≥56.384 dB,IR≥1.215 and ID≥9.184 dB.The area under the receiver operating characteristic curve were 0.590,0.877,0.815.There was no significant difference in the parameters of arterial phase,including peak time,initial growth time,rise time and the absolute value of peak intensity of lesions between the two groups(P>0.05).Multivariate analysis showed that the level of alphafetoprotein(AFP)and IR were risk factors for poor differentiation(P=0.001).CONCLUSION Among the parameters of Sonazoid-CEUS,IR in Kupffer phase exhibits superior diagnostic efficacy with high sensitivity and specificity in the diagnose of pathological differentiation of HCC.Combined with preoperative AFP level,a more accurate diagnosis will be obtained.Compared with portal vein phase,Kupffer phase showed the ability to identify HCC lesions more sensitive.These findings hold significant guiding implications and reference value for clinical practice.展开更多
目的:评估基于注射用全氟丁烷微球[商品名示卓安(Sonazoid)]超声造影Kupffer期的深度学习模型预测肝细胞癌(hepatocellular carcinoma,HCC)微血管侵犯(microvascular invasion,MVI)的效能,并将其与影像组学模型及临床模型进行比较。方法...目的:评估基于注射用全氟丁烷微球[商品名示卓安(Sonazoid)]超声造影Kupffer期的深度学习模型预测肝细胞癌(hepatocellular carcinoma,HCC)微血管侵犯(microvascular invasion,MVI)的效能,并将其与影像组学模型及临床模型进行比较。方法:回顾并纳入2020年7月—2022年9月于广西医科大学第一附属医院接受Sonazoid超声造影检查的146例原发性HCC患者,以7∶3随机划分为训练集102例和验证集44例。基于肿瘤感兴趣区,使用ResNet101模型通过迁移学习提取深度学习特征,使用PyRadiomics提取影像组学特征。采用Mann-Whitney U检验、最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)算法进行特征降维。LASSO回归用于构建深度学习模型和影像组学模型,同时还基于临床特征构建一个临床模型。采用受试者工作特征曲线的曲线下面积(area under the curve,AUC)、灵敏度、特异度和准确度评估模型的诊断效能。DeLong检验用于比较模型间的诊断效能。结果:在训练集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.931(0.880~0.981)、0.823(0.744~0.903)、0.719(0.614~0.824)。在验证集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.895(0.757~1.000)、0.711(0.514~0.909)、0.606(0.390~0.822)。DeLong检验表明在训练集和验证集中,深度学习模型的诊断效能均优于影像组学模型及临床模型(P<0.05)。单因素及多因素logistic回归分析示甲胎蛋白和巴塞罗那临床肝癌分期可作为HCC患者MVI的独立预测因子(P<0.01)。结论:基于Sonazoid超声造影Kupffer期的深度学习模型在预测HCC患者MVI方面表现出优异的性能,有望成为预测MVI的无创影像学生物标志物。展开更多
文摘BACKGROUND The treatment technology of liver cancer is progressing.In addition to traditional surgical resection,combined therapies of immunotherapy based on immune checkpoint inhibitors,chemotherapy,and transcatheter arterial chemoembolization for hepatocellular carcinoma are more and more widely used.Accurate preoperative diagnosis of liver cancer will provide important information for comprehensive treatment and prognosis evaluation of liver cancer.Sonazoidcontrast-enhanced ultrasound is not only helpful for the qualitative diagnosis of liver lesions,but also has great potential in the diagnosis of histological differentiation of liver cancer.AIM To assess the differentiation of hepatocellular carcinoma(HCC)by utilizing the parameters and imaging features of Sonazoid-contrast-enhanced ultrasound(CEUS).METHODS A retrospective analysis was conducted on the CEUS data of 239 lesions through case-control study.These patients received Sonazoid-CEUS within one week before surgery and were confirmed as HCC by postoperative pathology.Within the cases,patients were further categorized into well-differentiated and poorlydifferentiated group.Time-intensity curves of the region of interest in both arterial and Kupffer phases were generated,allowing for the acquisition of quantitative parameters to assess the diagnostic efficacy in distinguishing lesions between these two groups and determining an appropriate cut-off value.RESULTS Univariate analysis showed that the absolute value of enhancement intensity(EIAV),intensity ratio(IR)and intensity difference(ID)in Kupffer phase were statistically different between the groups with different degree(P=0.015,P=0.000,P=0.000).The sensitivity and specificity were 40.2%,82.4%,80.4% and 78.1%,86.9% and 74.5%,respectively,for differentiating HCC lesions with EIAV≥56.384 dB,IR≥1.215 and ID≥9.184 dB.The area under the receiver operating characteristic curve were 0.590,0.877,0.815.There was no significant difference in the parameters of arterial phase,including peak time,initial growth time,rise time and the absolute value of peak intensity of lesions between the two groups(P>0.05).Multivariate analysis showed that the level of alphafetoprotein(AFP)and IR were risk factors for poor differentiation(P=0.001).CONCLUSION Among the parameters of Sonazoid-CEUS,IR in Kupffer phase exhibits superior diagnostic efficacy with high sensitivity and specificity in the diagnose of pathological differentiation of HCC.Combined with preoperative AFP level,a more accurate diagnosis will be obtained.Compared with portal vein phase,Kupffer phase showed the ability to identify HCC lesions more sensitive.These findings hold significant guiding implications and reference value for clinical practice.
文摘目的:评估基于注射用全氟丁烷微球[商品名示卓安(Sonazoid)]超声造影Kupffer期的深度学习模型预测肝细胞癌(hepatocellular carcinoma,HCC)微血管侵犯(microvascular invasion,MVI)的效能,并将其与影像组学模型及临床模型进行比较。方法:回顾并纳入2020年7月—2022年9月于广西医科大学第一附属医院接受Sonazoid超声造影检查的146例原发性HCC患者,以7∶3随机划分为训练集102例和验证集44例。基于肿瘤感兴趣区,使用ResNet101模型通过迁移学习提取深度学习特征,使用PyRadiomics提取影像组学特征。采用Mann-Whitney U检验、最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)算法进行特征降维。LASSO回归用于构建深度学习模型和影像组学模型,同时还基于临床特征构建一个临床模型。采用受试者工作特征曲线的曲线下面积(area under the curve,AUC)、灵敏度、特异度和准确度评估模型的诊断效能。DeLong检验用于比较模型间的诊断效能。结果:在训练集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.931(0.880~0.981)、0.823(0.744~0.903)、0.719(0.614~0.824)。在验证集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.895(0.757~1.000)、0.711(0.514~0.909)、0.606(0.390~0.822)。DeLong检验表明在训练集和验证集中,深度学习模型的诊断效能均优于影像组学模型及临床模型(P<0.05)。单因素及多因素logistic回归分析示甲胎蛋白和巴塞罗那临床肝癌分期可作为HCC患者MVI的独立预测因子(P<0.01)。结论:基于Sonazoid超声造影Kupffer期的深度学习模型在预测HCC患者MVI方面表现出优异的性能,有望成为预测MVI的无创影像学生物标志物。
