摘要
目的对术前双期多模影像(增强CT和MRI)与实时双期C臂锥形束CT(DPCBCT)融合图像和单期单一影像技术的图像进行比较,评估多模融合影像对肝脏占位病灶检测能力的提高。方法选择肝脏占位病灶患者20例,其中男性19例,女性1例;年龄45~63岁,平均年龄54岁。经动脉化学治疗栓塞术(TACE)术前均接受肝脏增强CT或MRI扫描,得到双期图像。DPCBCT则在TACE术初期进行采集,完成采集后在控制室立即将动脉期和门静脉期图像与已有的双期图像分别进行融合,对融合后动脉期或门静脉期的肝脏占位病灶的多模图像并排显示,进行检测和评估。并分别计算单模影像和多模融合影像的检出率、漏诊率和误诊率。融合精度则在TACE术后进行回顾性评估。结果基于实时多模融合影像进行的病灶分析显示,确定116个动脉期和149个门静脉期病灶。术前单模影像(增强CT或MRI)的动脉期检出率、误诊率、漏诊率分别为87.9%、12.1%、0.0%,而DPCBCT则分别为95.7%、5.1%、0.9%;在门静脉期增强CT/MRI分别是77.2%、22.8%、0.0%,而门静脉期DPCBCT分别为97.3%、26.8%、0.0%。融合影像与单模影像相比具有更高的病灶检出率,尤其是较术前CT/MRI图像而言。DPCBCT对于病灶的检测更敏感,但相较增强CT/MRI具有更高的病灶误诊率。多模融合的精确度为95%。结论双期DPCBCT和增强CT/MRI多模融合是可行可靠的。该方法可以改善肝脏占位病灶的检测能力,以便更好地帮助施术者更精确地定位病灶,优化TACE治疗计划。
Objective To evaluate the improvement of hepatic focal lesion(HFL) detectobility by multi-modality image fusion, and compare preoperative dual-phase multimodal images(enhanced CT and MRI) with real-time dual-phase C-arm cone beam CT(DPCBCT) fusion images and single-mode single imaging techniques.Methods A total of 20 consecutive patients were enrolled,which included 19 males and 1 female,aged 45 - 63 years old with mean age of 54 years old.The liver enhanced CT or MRI scans were performed before transcatheter arterial chemotherapy embolization(TACE),then biphasic were obtained.The DPCBCT scan was performed at initial TACE,then arterial phase(AP) and venous phase(VP) images of DPCBCT were fused with biphasic MDCT/MRI images respectively,HFL was detected and evaluated alongside for each fused phase. The detection rate,missed diagnosis rate and misdiagnosis rate of single-mode images and multi-modality fusion images were calculated respectively.The fusion accuracy was evaluated retrospectively after TACE.Results The 116 arterial and 149 portal venous lesions were determined based on lesion analysis by real time multi-modality.The preoperative arterial phase detection rate,misdiagnosis rate and missed diagnosis rate of single-mode images(enhanced CT or MRI) were 87.9 %,12.1 %,and 0.0 %,respectively,and those rates of DPCBCT were 95.7 %,5.1 % and 0.9 %,respectively.For portal venous enhanced CT/MRI were 77.2 %,22.8 %,0.0 % and portal venous DPCBCT were 97.3 %,26.8 %,0.0 %.The fusion image had a higher detection rate than the single-mode image,especially than pre-operative CT/MRI images.However,it showed significantly higher detectability but more misdiagnosing lesions than CT/MRI.The accuracy of multi-modality fusion was 95 %.Conclusion It is demonstrated that the biphasic multi-modality fusion between DPCBCT and CT/MRI images is feasible,which could significantly improve the detectability of HFL and help better localization of HFL and optimize treatment strategy of TACE.
作者
曾帅
王嵇
程杰军
赵亮
管逊
马鋆
郭哲中
ZENG Shuai;WANG Ji;CHENG Jie-jun;ZHAO Liang;GUAN Xun;MA Jun;GUO Zhe-zhong(Department of Radiology,Renji Hospital,School of Medicine,Shanghai JiaoTong University,Shanghai 200127,China)
出处
《生物医学工程与临床》
CAS
2019年第4期447-451,共5页
Biomedical Engineering and Clinical Medicine
基金
上海市科委基金资助课题(14411968100)
