Influencing factors of quality in contrast-enhanced ultrasound quantitative perfusion analysis for hepatocellular carcinoma and a reproducibility study

Objective To investigate the influencing factors of quality in contrast-enhanced ultrasound (CEUS) perfusion analysis for hepatocellular carcinoma (HCC) and the reproducibility of using CEUS perfusion analysis on HCC. Methods A total of 136 HCC lesions were undergone CEUS perfusion analysis. Maximum intensity (Imax), rise time (RT), time to peak (TTP), mean transit time (mTT) and quality of fit (QOF) of HCC lesion, top-enhanced region inside HCC lesion and adjacent liver parenchyma were measured. The quality of perfusion analysis was classified into three grades Good (QOF>75%), Common [QOF (50%-75%)], and Poor (QOF<75%). Ultrasound machine, patients′ age, tumor depth, tumor size and tumor location were recorded and compared among the three groups. 20 consecutive patients with 20 HCC lesions were received CEUS perfusion analysis by two operators and another 20 consecutive HCC lesions were received CEUS perfusion analysis by one operator at different time. Intra-class correlation coefficient (ICC) was used to evaluate the reproducibility of inter- and intra- observer. Results Fifty-six HCC lesions were classified as Good , 39 as Common and 39 as Poor , respectively. Tumor size (P=0.015) and tumor location (P=0.041) were significantly different among the three groups. Tumor size >3.0 cm and tumor located in S3, S4b, S5, S6 were apt to gain a better QOF. No significant difference was found for the different ultrasound machine, tumor depth and patients′ age. For all CEUS perfusion analysis parameters of both lesion and adjacent parenchyma, intra-class correlation coefficient of inter-and intra-observer were higher than 0.90, the reproducibility of CEUS perfusion analysis was good. For Imax, RT and TTP in top-enhanced region inside HCC lesion, ICC of inter- and intra- observer was all higher than 0.90 (good). For mTT in highest enhanced area inside HCC lesion, ICC of inter- and intra-observer were 0.459 (poor), 0.609 (common), respectively. Conclusions The parameters of HCC quantitative perfusion analysis were reproducible in peripheral liver parenchyma and tumor, but the mTT repeatability was poor in the highest enhancement area of the tumor. The size and location of the lesion were the factors affecting quantitative analysis. The quantitative analysis of >3.0 cm and the tumor located in S3, S4b, S5, S6 in the liver were satisfactory. Key words: Carcinoma, hepatocellular/US; Phlebography; Reproducibility of results