分享:
分享到微信朋友圈
X
临床研究
基于术前增强MRI影像组学分析的列线图模型预测肝细胞癌切除术后复发风险的价值
王晴 盛晔 刘海峰 朱祖辉 邢伟

Cite this article as: Wang Q, Sheng Y, Liu HF, et al. Preoperative contrast-enhanced MRI based on radiomics analysis to predict the recurrence of hepatocellular carcinoma after resection[J]. Chin J Magn Reson Imaging, 2022, 13(12): 93-99.本文引用格式:王晴, 盛晔, 刘海峰, 等. 基于术前增强MRI影像组学分析的列线图模型预测肝细胞癌切除术后复发风险的价值[J]. 磁共振成像, 2022, 13(12): 93-99. DOI:10.12015/issn.1674-8034.2022.12.016.


[摘要] 目的 探讨术前增强MRI影像组学分析的列线图模型预测肝细胞癌(hepatocellular carcinoma, HCC)切除术后复发风险的价值。材料与方法 回顾性分析2015年8月至2020年8月在常州市第一人民医院进行HCC切除术的患者资料,共纳入164例于术前2周内进行增强MRI检查的患者病例,随机分为训练集(115例)和测试集(49例)。单因素及多因素Cox回归分析术前临床、病理及影像学特征与术后复发关系。运用最小绝对收缩和选择算子的Cox回归进行影像组学分析。联合影像组学标签及肿瘤复发的独立预测因素建立列线图预测模型,并在测试集进行验证。校准曲线观察模型预测概率与实际观察值一致性。根据训练集影像组学标签界值对HCC术后肿瘤复发风险分层,Kaplan-Meier法绘制生存曲线、Log-rank检验比较风险亚组间的生存差异。结果 肿瘤边界、肿瘤坏死、影像组学标签为预测HCC术后肿瘤复发独立因素(风险比分别为2.1、2.5、64.1,95%可信区间分别为1.3~3.3、1.5~4.3、20.6~199.9,P<0.05)。列线图模型预测肿瘤复发风险的C-index分别为训练集0.796(0.738~0.854)和测试集0.784(0.684~0.885)。模型预测概率与实际观察值有较好一致性。按影像组学标签界值进行复发风险分层,低危组无复发生存率较高,较高危组差异在训练集及测试集中均具有统计学意义(训练集:χ2=52.88,P<0.001;测试集:χ2=4.14,P=0.042)。结论 基于术前增强MRI影像组学分析的预测模型可有效预测HCC切除术后复发风险,有助于HCC术后患者个体化管理。
[Abstract] Objective To develop a preoperative MRI model based on radiomics analysis for predicting recurrence of hepatocellular carcinoma (HCC) patients after resection.Materials and Methods This retrospective study included 164 HCC patients (training set: n=115, testing set: n=49) who performed hepatectomy and preoperative gadoxetic acid-enhanced MRI within 2 weeks before resection between August 2015 and August 2020. The univariable and multivariable Cox regression analyses were performed to identify clinical-pathologic-radiologic factors associated with recurrence-free survival (RFS). The radiomics models were constructed using least absolute shrinkage and selection operator Cox regression. The combined nomogram model merging independent factors and radscore was built to predict the RFS of HCC patients after resection and the predictive performance of nomogram model was evaluated with C-index and calibration curves. Kaplan-Meier survival analysis was used to assess the association of the models with RFS.Results The combined nomogram model integrating the tumor margin [HR=2.1, 95% confidence interval (CI): 1.3 to 3.3], necrosis (HR=2.5, 95% CI: 1.5 to 4.3) and the radscore (HR=64.1, 95% CI: 20.6 to 199.9) showed good predictive efficacy for recurrence of HCC patients after resection with a C-index of 0.796 (0.738 to 0.854) in the training set and 0.784 (0.684 to 0.885) in the test set. Calibration curves demonstrated good agreement between model-predicted probabilities and observed outcomes. There was significant difference for recurrence rates between predicted low-risk group and high-risk group in the training set (χ2=52.88, P<0.001) and the test set (χ2=4.14, P=0.042).Conclusions The nomogram model demonstrated good performance for predicting recurrence of HCC patients after resection, thus may help personalized clinical management of HCC patients.
[关键词] 肝细胞肝癌;复发预测;磁共振成像;影像组学;列线图
[Keywords] hepatocellular carcinoma;recurrence;magnetic resonance imaging;radiomics;nomogram

王晴 1   盛晔 2   刘海峰 1   朱祖辉 1   邢伟 1*  

1 苏州大学附属第三医院(常州市第一人民医院)放射科,常州 213200

2 苏州大学附属第三医院(常州市第一人民医院)介入放射科,常州 213200

邢伟,E-mail:suzhxingwei@suda.edu.cn

作者利益冲突声明:全体作者均声明无利益冲突。


基金项目: 常州市卫健委青年基金 QN202111
收稿日期:2022-08-03
接受日期:2022-12-12
中图分类号:R445.2  R735.7 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2022.12.016
本文引用格式:王晴, 盛晔, 刘海峰, 等. 基于术前增强MRI影像组学分析的列线图模型预测肝细胞癌切除术后复发风险的价值[J]. 磁共振成像, 2022, 13(12): 93-99. DOI:10.12015/issn.1674-8034.2022.12.016.

       肝细胞癌(hepatocellular carcinoma, HCC)是世界第二致死的恶性肿瘤[1]。尽管HCC切除术是中国肝癌分期方案(China Liver Cancer Staging, CNLC)Ⅰa~Ⅱa期HCC的首选治疗方案,但是术后复发率较高,早期复发概率高达70%[2]。此外,HCC治疗手段众多,术前根据预测结果制订个体化选择治疗方案、优化患者术后随访及辅助治疗手段具有重要意义[3]。因此,有效预测HCC患者术后复发风险的手段值得探索。

       影像组学分析通过提取肿瘤内部影像特征参数构建影像组学模型[4, 5, 6, 7]。以往研究发现影像组学分析在HCC定性诊断、疗效评估、病理学特征评估及复发预测方面具有良好的诊断价值[8, 9, 10]。MRI对病灶诊断的敏感性优于计算机断层摄影(computer tomogram, CT)[11],基于MRI进行影像组学分析,结合临床资料构建模型可有效预测HCC切除术后复发风险[12, 13, 14]。但目前个体化预测HCC患者无复发生存时间(recurrence-free survival, RFS)的研究较少,且诊断效能C-index仅为0.607~0.725[15]。本研究旨在基于HCC患者术前增强MRI图像的影像组学分析与复发时间相关的独立因素,构建个体化预测HCC患者术后肿瘤复发风险的列线图模型。

1 材料与方法

       本研究经常州市第一人民医院伦理委员会批准,免除受试者知情同意,批准文号:(2022)教第039号。本研究符合《赫尔辛基宣言》对伦理学的要求。回顾性分析2015年8月至2020年8月常州市第一人民医院所有符合以下标准的病例资料。纳入标准:(1)进行肝脏部分切除术且经病理证实为HCC;(2)CNLC Ⅰa~Ⅱa期;(3)无肿瘤破裂;(4)未合并其他恶性肿瘤;(5)HCC瘤体完全切除且术后保留足够肝组织以维持肝功能;(6)肝癌切除术前两周内进行上腹部MRI增强扫描。排除标准:(1)患者手术前接受放疗、靶向治疗、介入治疗或肿瘤综合治疗的病例;(2)因呼吸、运动伪影或金属伪影致MRI图像质量降低影响术前HCC影像分析的病例;(3)临床病理资料缺失的病例。164例HCC患者病例纳入本研究,所有入组病例按照7∶3的比例随机分为训练集及测试集,训练集进行模型构建,测试集验证模型效能。

1.1 随访

       经门诊或电话随访。入组患者每3个月复查腹部CT或MRI。无复发生存时间(recurrence-free survival, RFS)定义为手术日至影像学检查发现明确新发病灶或经肝动脉造影或病理证实新发病灶的时间。

1.2 MRI检查

       检查设备采用德国Siemens Verio 3.0 T MRI系统和腹部32通道相控阵体线圈。禁食4~6 h后行上腹部MRI,患者取仰卧位,头先进,扫描范围为膈顶至肝下缘。扫描序列包括:横断面T1WI、T2WI、化学位移成像(正反相位T1WI)、肝脏三维容积多期动态增强成像(动脉期、门脉期、延迟期)。(1)T1WI序列扫描参数:TR 150~170 ms,TE同相位4.0 ms,矩阵320×192,视野38 cm×32 cm,层厚5 mm,层间距0 mm。(2)脂肪抑制T2WI序列扫描参数:TR 6059 ms,TE 80 ms,矩阵300×128,视野40 cm×38 cm,层厚5 mm,层间距0 mm。(3)肝脏三维容积多期动态增强成像:静脉给药Gd-DTPA(gadolinium-diethylenetriamine pentaacetic acid,马根维显,德国拜耳医药保健有限公司),剂量0.1 mmol/kg,速率2 mL/s,随后使用生理盐水20 mL冲洗;增强扫描(动脉早期10~20 s、动脉期20~30 s、门静脉期60~70 s、平衡期180 s)参数:TR 3.7 ms,TE 1.7 ms,矩阵320×192,视野38 cm×30 cm,层厚3 mm,层间距0 mm。

1.3 一般资料及影像学分析

       收集患者临床特征,包括年龄、性别、甲胎蛋白(alpha-fetoprotein, AFP)、CNLC分期;病理学特征包括病理分级、微血管浸润(microvascular invasion, MVI)、术后切缘、病理卫星灶等指标。

       由2名分别具有8年和10年腹部影像诊断经验的医师采用双盲法对图像进行独立影像学分析,当2名医师对影像特征判断结果不一致时,由具有30年影像诊断经验的主任医师进行最终判断。影像学分析特征包括肿瘤大小、肿瘤数量、肿瘤内坏死、动脉期强化、静脉期退出、肿瘤边界、假包膜、肝硬化背景、LI-RADS评分。肿瘤大小为病灶最大直径;多发病灶以最大病灶为研究对象;AFP取自然对数进行分析。

1.4 影像组学分析

       由10年腹部影像诊断经验的医生利用3D Slicer(version 4.9.0;http://www.slicer.org)于MRI多期增强扫描(动脉期、门脉期、延迟期)图像上以填充方式逐层勾画HCC全肿瘤感兴趣区域(region of interest, ROI)。所有图像按照1 mm×1 mm×1 mm体素重采样,通过开源PyRadiomics软件包(version 2.12;https://pyradiomics.readthedocs.io/en/2.1.2/)对图像进行特征提取。影像组学特征包括50个一阶特征,150纹理特征,750个小波变换特征,每位患者共提取2850个影像组学特征。所有特征按照Z-score进行标准化处理,最大相关最小冗余进行数据降维,聚类分析(cluster)进行特征筛选,采用最小绝对收缩和选择算法(least absolute shrinkage and selection operator, LASSO)-Cox算法联合10折交叉验证方法建模。

1.5 预测模型

       本研究在训练集中构建列线图。运用单因素及多因素Cox回归分析筛选与术后复发相关的临床、病理、影像学参数,依据赤池信息量准则采用逐步向后剔除变量法与影像组学标签构建Cox模型预测HCC切除术后复发的列线图。比较测试组C-index值评估模型预测能力。利用Bootstrap方法进行1000次重复抽样绘制校准曲线。根据训练集中影像学标签界值进行术后复发风险分层,将入组病例分为低危组、高危组[16]

1.6 统计学分析

       所有统计学分析基于R 3.4.4(http://www.r-project.org)和EmpowerState软件(www.empowerstats.com,X&Y_solutions, Inc. Boston MA)完成。计量资料采用均数±标准差(x¯±s)表示、符合正态分布的定量资料采用独立样本t检验,偏态分布的定量资料以中位值(四分位间距)表示、采用Mann-Whitney U检验,定性资料采用卡方检验、分类资料采用Fisher精确概率检验。采用Kaplan-Meier法绘制生存曲线、计算术后无复发生存率,并进行Log-rank检验比较两个风险亚组间的生存差异。P<0.05为差异有统计学意义。

2 结果

2.1 临床资料

       164例HCC术后患者,男130例,女34例,年龄(61.3±10.8)岁。训练集和测试集的临床、病理因素、影像特征组间差异均无统计学意义,详见表1。训练集随访时间为2~72个月,中位随访时间为20个月;测试集随访时间为2~66个月,中位随访时间为23月。其中96例随访过程中复发(训练集69例,测试集27例),68例未复发或删失(训练集46例,测试集22例),训练集2年复发率为58.9%,测试集2年复发率为50.4%,χ2=1.0975,P=0.294,差异不具有统计学意义。

表1  训练集与测试集肝细胞癌切除术后患者一般资料比较
Tab. 1  Comparison of the base line characteristics for HCC patients after resection between the training set and the test set

2.2 影像组学分析结果

       2850个影像组学特征中,最大相关最小冗余法进行数据降维、聚类分析筛选出排序前30位的影像组学特征。最终,LASSO-Cox回归分析纳入15个系数非0影像组学特征构建模型。公式如下:影像组学标签=0.00341×X1(PVP)-0.10451×X3(DP)-0.00781×X8(AP)+0.04618×X9(AP)+0.02254×X13(PVP)+0.0497×X15(AP)+0.03968×X16(PVP)-0.08712×X17(DP)+0.31825×X19(AP)-0.29185×X21(PVP)-0.09172×X25(DP)+0.125×X27(AP)+0.18879×X28(PVP)+0.13087×X29(AP)+0.01898×X30(PVP)。感兴趣区选择方法如图1及LASSO回归分析如图2所示。影像组学模型在训练集、测试集对预测HCC术后复发风险均具有一定价值,C-index分别为0.761(95%可信区间为0.700~0.822)和0.734(95%可信区间为0.635~0.833)。

图1  感兴趣区选择示意图。分别于动脉期、门脉期及延迟期沿肿瘤边界勾画全部病灶区并提取兴趣区内影像组学特征。
图2  影像组学模型构建。LASSO-Cox回归分析均方误差图(2A)及系数曲线图(2B)。黑色竖线定义为λ的最佳值,基于最小似然偏差与10折交叉验证方法选择最佳权重参数λ=0.056和Log(λ)=-2.886,最终筛选15个系数非0影像组学特征构建模型。
Fig. 1  Schematic diagram of region of interest (ROI) selection. ROI of whole tumor were automatically outlined around the lesion in magnetic resonance imaging (arterial phase, portal phase and delay phase), followed by extractions of radiomic features including shape, intensity, texture features.
Fig. 2  Construction of radiomics model. 2A-2B: We selected λ values of 0.056 [Log(λ)=-2.886] in the LASSO model using a 10-fold cross-validation approach to identify the optimal subsets of radiomics features. Finally, 15 nonzero coefficients were included in the model.

2.3 预测模型构建及列线图

       单因素Cox回归分析显示肿瘤数量、CNLC、肿瘤大小、肿瘤坏死、肿瘤边界是影响肝细胞癌切除术后肿瘤复发的相关因素(P<0.05);多因素Cox回归分析显示HCC术后复发预测模型由肿瘤边界、肿瘤坏死、影像组学标签构成(P<0.05),详见表2。风险预测方程为Y=0.715×肿瘤边界(0:平滑;1:不平滑)+0.757×肿瘤坏死(0:无坏死;1:有坏死)+5.068×影像组学标签,构建的列线图见图3A。训练集及测试集中,该模型预测HCC切除术后复发风险的诊断效能良好,C-index分别为0.796(95%可信区间为0.738~0.854)和0.784(95%可信区间为0.684~0.885),详见表3。校准曲线提示影像组学模型预测的复发概率与实际观察值具有较好一致性(图3B)。

图3  肝癌切除术后复发预测模型的列线图(3A)及预测模型的校准曲线(3B)。校准曲线提示影像组学模型的预测概率与实际观察值具有较好一致性。
图4  根据预测风险模型定义的复发风险进行分层,Kaplan-Meier生存分析RFS。4A:训练集;4B:测试集。
Fig. 3  A combined nomogram was developed in the training set (3A). Calibration curves to show the calibration of the combined nomogram in terms of the agreement between the predicted and the observed RFS in the training set (3B).
Fig. 4  Kaplan–Meier survival analysis of recurrence-free survival time according to risk strata defined by radiomics model. 4A: training dataset; 4B: validation dataset.
表2  肝切除术后肿瘤复发的单因素分析及多因素分析
Tab. 2  Univariable and multivariate analysis of baseline characteristics for recurrence-free survival evaluation of hepatocellular carcinoma patients after resection
表3  影像组学模型及风险预测模型对肝癌切除术后复发风险预测效能
Tab. 3  Performance of the radiomics model and combined model for predicting the recurrence risk of hepatocellular carcinoma after resection

2.4 复发风险评估

       按照训练集中影像学标签的界值(影像学标签=0.475)进行分层。小于界值归为低危组(n=78),大于界值归为高危组(n=86)。Kaplan-Meier检验结果显示(图4):在训练集及测试集,低危组内肿瘤术后复发率均低于高危组,且Log-rank检验分析显示在低危组及高危组间RFS差异有统计学意义(训练集:χ2=52.88,P<0.001;测试集:χ2=4.14,P=0.042)。训练集中位RFS分别为57个月(低危组)和9个月(高危组),测试集中位RFS分别为30个月(低危组)和13个月(高危组)。训练集中,低危组的HCC术后1年、5年肿瘤复发率为11%、47%,高危组1年、5年复发率为60%、99%;测试集中低危组的HCC术后1年、5年肿瘤复发率分别为21%、69%,高危组的HCC术后1年、5年肿瘤复发率分别41%、98%。

3 讨论

       本研究基于MRI增强图像构建影像组学标签,联合临床、影像学定性特征、病理特征构建模型预测HCC切除术后复发风险。结果表明影像组学标签、肿瘤内部坏死及肿瘤边界是预测HCC切除术后复发的独立因素,且该复发预测模型具有良好的效能及模型校准度,可能成为HCC切除患者复发预测的术前评估手段,为HCC患者个体化手术及后续治疗方案提供依据。

3.1 基于术前增强MRI图像影像组学分析的列线图模型预测HCC切除患者复发的价值

       以往研究证实影像组学分析对HCC术前诊断、疗效评估及治疗方案选择具有良好的价值[17, 18, 19, 20]。Wang等[7]构建影像组学模型鉴别HCC、胆管细胞癌及混合癌效能高达0.91。Sun等[21]探索不可切除HCC经导管动脉栓塞治疗疗效的预测效能为0.80。多个研究[10,22, 23, 24]针对预测MVI的问题构建影像组学预测模型,效能为0.668~0.879不等,且影像组学预测模型对HCC免疫评分评估(CD3+和CD8+T细胞的密度)亦有价值(曲线下面积分别为0.823、0.904)[25]

       HCC术后发生肝内早期复发的患者五年生存率降低24%,中位生存期仅为54个月[26],因此术前预测HCC术后复发风险并进行个体化术后治疗方案管理具有重要临床意义。以往研究表明影像组学预测模型对HCC切除术后2年内复发的效能约为0.654~0.785[26],HCC术后5年复发率的预测效能为0.776~0.780[27]。同时基于影像组学的对比研究亦发现:MRI在诊断小肝癌、预测及鉴别肿瘤良恶性及预测MVI方面均取得良好效果,且价值优于CT[11,22,28]。因此,本研究基于增强MRI进行影像组学分析。以往定义2年内复发为早期复发、5年内复发为晚期复发可能会一定程度造成随访复发时间的信息丢失,HCC术后患者1年和5年无复发生存率较低,仅为88.9%和56.2%[11],不利于HCC患者的个体化预测。因此,预测患者首次复发时间对HCC患者术后的个体化治疗尤其重要。本研究基于MRI影像组学分析构建模型有效预测HCC切除术后首次复发时间,效能达到0.784~0.796,略高于Liu等[29]及Zhang等[15]的研究效能(C-index为0.725~0.741),略低于Ji等[3]的结果(C-index为0.733~0.801)。除影像组学标签,本研究提示肿瘤边界及肿瘤内部坏死均为预测肝癌复发风险的独立因子,这与以往研究结果部分相仿[3,30]。肿瘤结节外生长导致肿瘤边界模糊、分界不清,是MVI出现的重要标志;同时肿瘤边界不光滑是门静脉侵犯和肝内转移的重要标志。因此,由于肿瘤内异质性及生长各向异性,肿瘤边界不光滑多提示肿瘤周围的肝组织受浸润,肿瘤恶性程度较高,与HCC患者复发密切相关[31]。肿瘤内部坏死是实体肿瘤的常见特征,肿瘤内发生自发坏死的原因可能与宿主相关条件下产生的炎症反应及与宿主乙肝病毒的水平有关,以往HCC研究发现HCC瘤内坏死率大于50%提示整体生存率更低、预后更差,肿瘤内部自发坏死与患者的总生存时间和无复发生存时间的降低有关[32],这与本研究结果一致。因此,本研究认为肿瘤浸润型边界、瘤内坏死是HCC患者复发的危险因素。尽管肿瘤大小、肿瘤数量、CNLC与肝癌术后复发相关性在单因素分析中具有显著性,但多因素分析无统计学意义。可能由于较放射组学标签,肿瘤大小、肿瘤数量对预测模型影响权重较小,或与入组患者选择偏倚有关。肿瘤大小、肿瘤数量与手术方式、残留肝体积大小密切相关,与患者术后免疫应答有关[33],因此需要更大样本、多中心的研究对此特征进行分析。

3.2 术前预测HCC术后复发风险对治疗决策具有重要意义

       本研究中预测为复发高危且随访证实发生早期复发的63例患者(63/164),如尽早改用肝移植或早期进行辅助肝动脉化疗栓塞术(transcatheter arterial chemoembolization, TACE)治疗、综合治疗可能更有助于患者生存率提高[34]。以往研究发现切除术后残留肝组织免疫抑制,癌前病变(如不典型再生结节)快速增殖易发生恶变,且该时期增殖细胞对TACE治疗十分敏感,因此对HCC切除术后复发高危的患者术后1个月内进行辅助性TACE治疗,可极大地减少肿瘤负荷,提高患者总生存期及无瘤生存时间[35]。但是,以往认定高复发风险患者标准不一[33],且主要依赖术后病理学特征,从而无法针对患者制订个体化的手术、治疗方案。因此,术前个体化预测HCC术后患者的复发风险,制订个体化的术中、术后管理方案具有重要的临床价值。本研究的HCC复发风险预测模型有望为患者个体化治疗方案选择提供依据。

3.3 本研究的局限性

       本研究尚有一定局限性:(1)本研究资料来自单中心,且样本量较小,模型的预测效能还需在多中心HCC患者中进一步验证;(2)本研究患者未注射肝特异性对比剂钆塞酸二钠(gadolinium ethoxybenzyl-diethylenetriaminepenta-acetic acid, Gd-EOB-DTPA),而术前Gd-EOB-DTPA增强MRI对HCC根治术后早期复发预测具有一定的价值[36],因此,需要进一步前瞻性研究进行完善;(3)本文为回顾性研究且部分患者随访时间相对较短,对远期复发及生存时间的预测效能有限。

       综上所述,本研究基于术前MRI增强图像的影像组学特征及临床资料构建HCC切除术后肿瘤复发的预测模型,实现术前预测HCC患者切除术后的复发风险,为HCC患者个体化治疗方案选择、术后管理提供依据。

[1]
Lee HA, Lee YS, Kim BK, et al. Change in the recurrence pattern and predictors over time after complete cure of hepatocellular carcinoma[J]. Gut Liver, 2021, 15(3): 420-429. DOI: 10.5009/gnl20101.
[2]
Dimitroulis D, Damaskos C, Valsami S, et al. From diagnosis to treatment of hepatocellular carcinoma: An epidemic problem for both developed and developing world[J/OL]. World J Gastroenterol, 2017, 23(29): 5282-5294 [2022-02-08]. https://www.wjgnet.com/1007-9327/full/v23/i29/5282.htm. DOI: 10.3748/wjg.v23.i29.5282.
[3]
Ji GW, Zhu FP, Xu Q, et al. Machine-learning analysis of contrast-enhanced CT radiomics predicts recurrence of hepatocellular carcinoma after resection: a multi-institutional study[J]. EBioMedicine, 2019, 50: 156-165. DOI: 10.1016/j.ebiom.2019.10.057.
[4]
Nardone V, Reginelli A, Grassi R, et al. Delta radiomics: a systematic review[J]. Radiol Med, 2021, 126(12): 1571-1583. DOI: 10.1007/s11547-021-01436-7.
[5]
Chetan MR, Gleeson FV. Radiomics in predicting treatment response in non-small-cell lung cancer: current status, challenges and future perspectives[J]. Eur Radiol, 2021, 31(2): 1049-1058. DOI: 10.1007/s00330-020-07141-9.
[6]
Wakabayashi T, Ouhmich F, Gonzalez-Cabrera C, et al. Radiomics in hepatocellular carcinoma: a quantitative review[J]. Hepatol Int, 2019, 13(5): 546-559. DOI: 10.1007/s12072-019-09973-0.
[7]
Wang X, Wang S, Yin X, et al. MRI-based radiomics distinguish different pathological types of hepatocellular carcinoma[J/OL]. Comput Biol Med, 2022, 141: 105058 [2022-02-08]. https://linkinghub.elsevier.com/retrieve/pii/S0010-4825(21)00852-0. DOI: 10.1016/j.compbiomed.2021.105058.
[8]
Harding-Theobald E, Louissaint J, Maraj B, et al. Systematic review: radiomics for the diagnosis and prognosis of hepatocellular carcinoma[J]. Aliment Pharmacol Ther, 2021, 54(7): 890-901. DOI: 10.1111/apt.16563.
[9]
Fontaine P, Riet FG, Castelli J, et al. Comparison of feature selection in radiomics for the prediction of overall survival after radiotherapy for hepatocellular carcinoma[C]//2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society. Montreal, QC, Canada. IEEE, 2020: 1667-1670. DOI: 10.1109/EMBC44109.2020.9176724.
[10]
Chong HH, Yang L, Sheng RF, et al. Multi-scale and multi-parametric radiomics of gadoxetate disodium-enhanced MRI predicts microvascular invasion and outcome in patients with solitary hepatocellular carcinoma≤5 cm[J]. Eur Radiol, 2021, 31(7): 4824-4838. DOI: 10.1007/s00330-020-07601-2.
[11]
Wang GB, Zhu SC, Li XK. Comparison of values of CT and MRI imaging in the diagnosis of hepatocellular carcinoma and analysis of prognostic factors[J]. Oncol Lett, 2019, 17(1): 1184-1188. DOI: 10.3892/ol.2018.9690.
[12]
Gao WY, Wang WT, Song DJ, et al. A predictive model integrating deep and radiomics features based on gadobenate dimeglumine-enhanced MRI for postoperative early recurrence of hepatocellular carcinoma[J]. Radiol Med, 2022, 127(3): 259-271. DOI: 10.1007/s11547-021-01445-6.
[13]
Chong HH, Gong YD, Pan XP, et al. Peritumoral dilation radiomics of gadoxetate disodium-enhanced MRI excellently predicts early recurrence of hepatocellular carcinoma without macrovascular invasion after hepatectomy[J]. J Hepatocell Carcinoma, 2021, 8: 545-563. DOI: 10.2147/JHC.S309570.
[14]
Zhao Y, Wu JJ, Zhang QH, et al. Radiomics analysis based on multiparametric MRI for predicting early recurrence in hepatocellular carcinoma after partial hepatectomy[J]. J Magn Reson Imaging, 2021, 53(4): 1066-1079. DOI: 10.1002/jmri.27424.
[15]
Zhang L, Hu JM, Hou JY, et al. Radiomics-based model using gadoxetic acid disodium-enhanced MR images: associations with recurrence-free survival of patients with hepatocellular carcinoma treated by surgical resection[J]. Abdom Radiol (NY), 2021, 46(8): 3845-3854. DOI: 10.1007/s00261-021-03034-7.
[16]
Fang SJ, Lai LQ, Zhu JY, et al. A Radiomics Signature-Based Nomogram to Predict the Progression-Free Survival of Patients With Hepatocellular Carcinoma After Transcatheter Arterial Chemoembolization Plus Radiofrequency Ablation[J/OL]. Front Mol Biosci, 2021, 8: 662366 [2022-02-08]. https://www.frontiersin.org/articles/10.3389/fmolb.2021.662366/full. DOI: 10.3389/fmolb.2021.662366.
[17]
Mokrane FZ, Lu L, Vavasseur A, et al. Radiomics machine-learning signature for diagnosis of hepatocellular carcinoma in cirrhotic patients with indeterminate liver nodules[J]. Eur Radiol, 2020, 30(1): 558-570. DOI: 10.1007/s00330-019-06347-w.
[18]
Liu QP, Yang KL, Xu X, et al. Radiomics analysis of pretreatment MRI in predicting tumor response and outcome in hepatocellular carcinoma with transarterial chemoembolization: a two-center collaborative study[J]. Abdom Radiol (NY), 2022, 47(2): 651-663. DOI: 10.1007/s00261-021-03375-3.
[19]
Jin ZC, Chen L, Zhong BY, et al. Machine-learning analysis of contrast-enhanced computed tomography radiomics predicts patients with hepatocellular carcinoma who are unsuitable for initial transarterial chemoembolization monotherapy: A multicenter study[J/OL]. Transl Oncol, 2021, 14(4): 101034 [2022-03-10]. https://linkinghub.elsevier.com/retrieve/pii/S1936-5233(21)00026-7. DOI: 10.1016/j.tranon.2021.101034.
[20]
Shen JX, Zhou Q, Chen ZH, et al. Longitudinal radiomics algorithm of posttreatment computed tomography images for early detecting recurrence of hepatocellular carcinoma after resection or ablation[J/OL]. Transl Oncol, 2021, 14(1): 100866 [2022-02-12]. https://linkinghub.elsevier.com/retrieve/pii/S1936-5233(20)30358-2. DOI: 10.1016/j.tranon.2020.100866.
[21]
Sun YJ, Bai HL, Xia W, et al. Predicting the outcome of transcatheter arterial embolization therapy for unresectable hepatocellular carcinoma based on radiomics of preoperative multiparameter MRI[J]. J Magn Reson Imaging, 2020, 52(4): 1083-1090. DOI: 10.1002/jmri.27143.
[22]
Meng XP, Wang YC, Zhou JY, et al. Comparison of MRI and CT for the prediction of microvascular invasion in solitary hepatocellular carcinoma based on a non-radiomics and radiomics method: which imaging modality is better?[J]. J Magn Reson Imaging, 2021, 54(2): 526-536. DOI: 10.1002/jmri.27575.
[23]
Xu X, Zhang HL, Liu QP, et al. Radiomic analysis of contrast-enhanced CT predicts microvascular invasion and outcome in hepatocellular carcinoma[J]. J Hepatol, 2019, 70(6): 1133-1144. DOI: 10.1016/j.jhep.2019.02.023.
[24]
田雅琪, 胡亚彬, 彭琪琪, 等. 钆塞酸二钠增强MRI列线图模型预测小肝癌微血管侵犯的价值[J]. 磁共振成像, 2021, 12(10): 57-60, 65. DOI: 10.12015/issn.1674-8034.2021.10.013.
Tian YQ, Hu YB, Peng QQ, et al. The value of Gd-EOB-DTPA-enhanced MRI nomogram model in predicting microvascular invasion of small solitary hepatocellular carcinoma[J]. Chin J Magn Reson Imaging, 2021, 12(10): 57-60, 65. DOI: 10.12015/issn.1674-8034.2021.10.013.
[25]
Chen SL, Feng ST, Wei JW, et al. Pretreatment prediction of immunoscore in hepatocellular cancer: a radiomics-based clinical model based on Gd-EOB-DTPA-enhanced MRI imaging[J]. Eur Radiol, 2019, 29(8): 4177-4187. DOI: 10.1007/s00330-018-5986-x.
[26]
Zhu HB, Zheng ZY, Zhao H, et al. Radiomics-based nomogram using CT imaging for noninvasive preoperative prediction of early recurrence in patients with hepatocellular carcinoma[J]. Diagn Interv Radiol, 2020, 26(5): 411-419. DOI: 10.5152/dir.2020.19623.
[27]
Li PX, Wu L, Li ZH, et al. Spleen Radiomics Signature: A Potential Biomarker for Prediction of Early and Late Recurrences of Hepatocellular Carcinoma After Resection[J/OL]. Front Oncol, 2021, 11: 716849 [2022-06-08]. https://www.frontiersin.org/articles/10.3389/fonc.2021.716849/full. DOI: 10.3389/fonc.2021.716849.
[28]
Liang WJ, Shao JY, Liu WH, et al. Differentiating Hepatic Epithelioid Angiomyolipoma From Hepatocellular Carcinoma and Focal Nodular Hyperplasia via Radiomics Models[J/OL]. Front Oncol, 2020, 10: 564307 [2022-02-08]. https://www.frontiersin.org/articles/10.3389/fonc.2020.564307/full. DOI: 10.3389/fonc.2020.564307.
[29]
Liu F, Liu D, Wang K, et al. Deep learning radiomics based on contrast-enhanced ultrasound might optimize curative treatments for very-early or early-stage hepatocellular carcinoma patients[J]. Liver Cancer, 2020, 9(4): 397-413. DOI: 10.1159/000505694.
[30]
Zhang Z, Jiang HY, Chen J, et al. Hepatocellular carcinoma: radiomics nomogram on gadoxetic acid-enhanced MR imaging for early postoperative recurrence prediction[J/OL]. Cancer Imaging, 2019, 19(1): 22 [2022-06-08]. https://cancerimagingjournal.biomedcentral.com/articles/10.1186/s40644-019-0209-5. DOI: 10.1186/s40644-019-0209-5.
[31]
Song L, Li J, Luo Y. The importance of a nonsmooth tumor margin and incomplete tumor capsule in predicting HCC microvascular invasion on preoperative imaging examination: a systematic review and meta-analysis[J]. Clin Imaging, 2021, 76: 77-82. DOI: 10.1016/j.clinimag.2020.11.057.
[32]
Bijelic L, Rubio ER. Tumor necrosis in hepatocellular carcinoma—unfairly overlooked?[J]. Ann Surg Oncol, 2021, 28(2): 600-601. DOI: 10.1245/s10434-020-09402-9.
[33]
Xie H, Tian ST, Cui L, et al. Adjuvant trans-arterial chemoembolization after hepatectomy significantly improves the prognosis of low-risk patients with R0-stage hepatocellular carcinoma[J]. Cancer Manag Res, 2019, 11: 4065-4073. DOI: 10.2147/CMAR.S195485.
[34]
季顾惟, 王科, 吴晓峰, 等. 基于CT检查影像组学早期肝细胞癌切除术后肿瘤复发的预测模型构建及其应用价值[J]. 中华消化外科杂志, 2020(2): 204-216. DOI: 10.3760/cma.j.issn.1673-9752.2020.02.014.
Ji GW, Wang K, Wu XF, et al. Construction and application value of CT-based radiomics model for predicting recurrence of early-stage hepatocellular carcinoma after resection[J]. Chin J Dig Surg, 2020(2): 204-216. DOI: 10.3760/cma.j.issn.1673-9752.2020.02.014.
[35]
Huo YR, Chan MV, Chan C. Resection plus post-operative adjuvant transcatheter arterial chemoembolization (TACE) compared with resection alone for hepatocellular carcinoma: a systematic review and Meta-analysis[J]. Cardiovasc Intervent Radiol, 2020, 43(4): 572-586. DOI: 10.1007/s00270-019-02392-6.
[36]
赵其煜, 戚元刚, 郭守芳, 等. Gd-EOB-DTPA增强MRI对肝细胞癌术后早期复发的预测价值[J]. 磁共振成像, 2021, 12(12): 18-23. DOI: 10.12015/issn.1674-8034.2021.12.004.
Zhao QY, Qi YG, Guo SF, et al. The value of Gd-EOB-DTPA enhanced magnetic resonance imaging for predicting early recurrence of hepatocellular carcinoma after resection[J]. Chin J Magn Reson Imaging, 2021, 12(12): 18-23. DOI: 10.12015/issn.1674-8034.2021.12.004.

上一篇 R2*图纹理分析预测肝细胞癌肝切除术后早期复发的价值
下一篇 三维酰胺质子转移成像鉴别前列腺癌伴骨转移与不伴骨转移的可行性研究
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2