分享:
分享到微信朋友圈
X
临床研究
基于磁共振动态增强图像直方图分析对肝癌TACE术后早期疗效评估及预测研究
王东东 李晓茗 杨柳 张斯佳

Cite this article as: WANG D D, LI X M, YANG L, et al. Evaluation and prediction of therapeutic effect of hepatocellular carcinoma after TACE based on contrast enhanced MRI histogram analysis[J]. Chin J Magn Reson Imaging, 2024, 15(1): 113-118, 131.本文引用格式:王东东, 李晓茗, 杨柳, 等. 基于磁共振动态增强图像直方图分析对肝癌TACE术后早期疗效评估及预测研究[J]. 磁共振成像, 2024, 15(1): 113-118, 131. DOI:10.12015/issn.1674-8034.2024.01.018.


[摘要] 目的 探讨基于对比增强磁共振成像(contrast-enhanced magnetic resonance imaging, CE-MRI)图像直方图参数对肝细胞癌(hepatocellular carcinoma, HCC)经导管肝动脉化疗栓塞(transcatheter arterial chemoembolization, TACE)术后早期疗效的评估和预测价值。材料与方法 回顾性分析我院2018年6月至2021年12月经病理活检证实或临床确诊为HCC的64例患者的影像及临床资料。所有患者均在TACE治疗前1个月内、首次TACE术后4~6周内行上腹部MR常规扫描及T1WI三期对比增强序列扫描,根据修正后实体瘤疗效评价标准(modified response evaluation criteria in solid tumors, mRECIST)对治疗后肿瘤疗效情况经随访将患者分为有效组34例和无效组30例。选用轴位T1WI对比增强(动脉期、门脉期、延迟期)HCC最大层面图像,通过MaZda软件人工手动勾画ROI提取病灶的直方图特征,包括均值、方差、偏度、峰度以及1、10、50、90和99百分位数值(分别记录为Perc1、Perc10、Pere50、Perc90和Perc99)共9个特征参数。采用SPSS软件分别比较TACE治疗前后有效组、无效组之间各参数的差异。对于评估肝癌术后疗效有统计学意义的参数,采用受试者工作特征(receiver operating characteristic, ROC)曲线评价其诊断效能,确定有预测价值的参数值及阈值。结果 TACE治疗前两组间差异有统计学意义的参数包括动脉期均值、Perc50、Perc90、Perc99和静脉期Perc1、延迟期峰度值(P<0.05);TACE治疗后两组间差异有统计学意义的参数包括延迟期均值、Perc50、Perc99值(P<0.05);其余各参数TACE治疗前后两组间比较差异无统计学意义(P>0.05)。ROC曲线显示各参数的效能为:术前动脉期均值>术后延迟期Perc99>术前动脉期Perc90>术后延迟期均值>术前延迟期峰度>术后延迟期Perc50>术前延迟期Perc50>术前动脉期Perc99,TACE术前动脉期均值的ROC曲线下面积最大[0.710(95% CI:0.583~0.817)]。结论 CE-MRI图像的直方图特征能够有效评估及预测肝癌TACE术后早期疗效,有助于指导临床治疗方案。
[Abstract] Objective To evaluate and predict the early curative effect of transcatheter arterial chemoembolization (TACE) for hepatocellular carcinoma based on contrast-enhanced magnetic resonance imaging (CE-MRI) histogram parameters of MRI.Materials and Methods Sixty-four patients with hepatocellular carcinoma confirmed by pathological biopsy or clinically diagnosed from June 2018 to December 2021 were selected for study. All patients underwent routine MR scan and multi-phase dynamic enhanced T1WI contrast-enhanced scan within 1 month before TACE treatment and 4-6 weeks after the first TACE treatment. According to the modified solid tumor efficacy evaluation criteria (mRECIST), the patients were divided into effective group of 34 cases and ineffective group of 30 cases. The HCC maximum axial T1WI images (arterial phase, portal phase and delayed phase) of HCC maximum slice images were selected for all patients, and the histogram features of the lesions were manually extracted by sketching ROI with MaZda software, including mean, variance, skewness, kurtosis and the values of the 1st, 10th, 50th, 90th and 99th percentiles (recorded as Perc1, Perc10, Pere50, Perc90 and Perc99, respectively). Statistical software was used to compare the differences of histogram parameters between effective group and ineffective group before and after TACE treatment. The parameters with statistical differences in the postoperative efficacy of HCC were evaluated by the receiver operating characteristic (ROC) curve, and the valuable parameter values and thresholds were obtained.Results The parameters with statistically significant differences between the two groups before TACE treatment including mean value, Perc50, Perc90, Perc99 in arterial phase, Perc1 in venous phase, and peak value in delayed phase (P<0.05); the parameters with statistically significant differences between the two groups after TACE treatment including the mean value of delay period, Perc50, Perc99 (P<0.05); there was no statistically significant difference in the other parameters between the two groups before and after TACE treatment (P>0.05). The ROC curve results showed that the efficacy of each parameter was as follows: preTACE arterial phase mean>postTACE delay phase Perc99>preTACE arterial phase Perc90>postTACE delay phase mean>preTACE delay phase kurtosis>postTACE delay phase Perc50>pretTACE delay phase Perc50>preTACE arterial phase Perc99. The area under the ROC curve of preTACE arterial phase mean was the highest, with a value of 0.710 (95% CI: 0.583-0.817).Conclusions The histogram features of CE-MRI can effectively evaluate and predict the characteristics of early curative effect after TACE of hepatocellular carcinoma, which helps guide clinical treatment plans.
[关键词] 肝细胞癌;磁共振成像;对比增强磁共振成像;直方图分析;纹理分析
[Keywords] hepatocellular carcinoma;magnetic resonance imaging;contrast enhanced magnetic resonance imaging;histogram analysis;texture analysis

王东东    李晓茗    杨柳    张斯佳 *  

河南中医药大学第五临床医学院(郑州人民医院)医学影像科,郑州 450053

通信作者:张斯佳,E-mail:13703923296@163.com

作者贡献声明::张斯佳设计本研究的方案,对稿件重要的内容进行了修改,获得了河南省医学科技攻关计划项目(编号:2018020831)资助;王东东起草和撰写稿件,获取、分析和解释本研究的数据,并获得了河南省医学科技攻关计划项目(编号:LHGJ20210708)资助;李晓茗、杨柳获取、分析和解释本研究的数据,对稿件重要的内容进行了修改;全体作者都同意发表最后的修改稿,同意对本研究的所有方面负责,确保本研究的准确性和诚信。


基金项目: 河南省医学科技攻关计划项目 LHGJ20210708,2018020831
收稿日期:2023-06-07
接受日期:2023-12-26
中图分类号:R445.2  R735.7 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2024.01.018
本文引用格式:王东东, 李晓茗, 杨柳, 等. 基于磁共振动态增强图像直方图分析对肝癌TACE术后早期疗效评估及预测研究[J]. 磁共振成像, 2024, 15(1): 113-118, 131. DOI:10.12015/issn.1674-8034.2024.01.018.

0 引言

       原发性肝癌(primary liver cancer, PLC)是中国最常见的恶性肿瘤之一,为世界第三大癌症死亡原因[1, 2],肝细胞癌(hepatocellular carcinoma, HCC)为其主要类型,由于HCC起病隐匿、早期无明显临床症状,大部分HCC患者确诊时已处于中晚期,失去手术机会。经导管肝动脉化疗栓塞(transcatheter arterial chemoembolization, TACE)治疗已广泛用于不能切除的中晚期HCC患者[3],是中晚期肝癌患者的首选治疗方案[4],它可以阻断肿瘤血供,使肿瘤缺血坏死[5],但TACE治疗后的肿瘤坏死率并不是100%,有时需反复治疗才能达到理想效果[6, 7]。因此,在TACE实施前后准确评估肿瘤活性对于患者预后是非常重要的,但目前尚缺乏对TACE术后疗效评估及预测的有效方法。近年来,纹理分析技术在医学图像分析中得到了越来越广泛的应用,通过对图像中病灶形态的勾画、特征的提取,进而获取一系列反映病灶本质特征的相关纹理参数,其中,一阶测量获取的纹理特征即直方图特征,反映了感兴趣区(region of interest, ROI)内的灰度频度分布[8, 9, 10],客观地反映了疾病尤其是肿瘤内部信息[11, 12, 13],可评估肿瘤的异质性和病理分级[14, 15, 16],目前已经越来越多地应用于肿瘤良恶性的鉴别[11, 12]及评估肿瘤治疗的疗效和预测[17, 18, 19, 20]。有学者基于术前MRI的纹理分析对肝癌TACE术后早期疗效进行预测[21],但只提取了动脉期及静脉期图像的纹理参数。本研究纳入了延迟期图像,通过对TACE术前后三期(动脉期、静脉期、延迟期)增强图像的纹理分析评估TACE术后疗效,旨在探讨基于磁共振T1WI三期增强图像直方图特征(一阶纹理参数)对肝癌TACE术后早期疗效的评估和预测价值。

1 材料与方法

1.1 研究对象

       本研究遵守《赫尔辛基宣言》,经我院伦理委员会批准,免除受试者知情同意,批准文号:2023011103。收集2018年6月至2021年12月在我院经TACE治疗的HCC患者。纳入标准:(1)所有患者TACE术前经病理证实为HCC或依据《原发性肝癌诊疗规范(2019年版)》[22]临床诊断为HCC [影像学分类达到肝脏影像报告和数据系统(Liver Imaging Reporting And Data System, LI-RADS)5类)];(2)肝功能Child-Pugh A或B级;(3)肝脏肿瘤数目≤2个,病灶最大直径>1 cm,且<10 cm。排除标准:(1)图像伪影较大,影响病灶勾画;(2)有放化疗病史或激素治疗史;(3)患者未到本研究随访终点退出者。所有入组患者均在术前1个月内及首次TACE治疗后4~6周内行MR常规扫描[T1WI、T2WI、扩散加权成像(diffusion weighted imaging, DWI)]和T1WI三期增强序列扫描(动脉期、门脉期、延迟期),后期随访扫描序列同上。

1.2 TACE治疗

       使用改良Seldinger法穿刺右侧股动脉,经微导管进入肝脏肿瘤供血动脉,根据肿瘤大小、部位和数量,经微导管缓慢注入表柔比星、碘化油混合物栓塞肿瘤,栓塞以后造影复查,直至肝肿瘤供血动脉消失,术后密切观察有无并发症。

1.3 疗效评估及分组

       所有入组患者首次TACE术后至少随访半年,随访终点复查上腹部MRI平扫及T1WI三期增强检查。按照修正后实体瘤疗效评价标准(modified response evaluation criteria in solid tumors, mRECIST)[23]对TACE术后早期疗效的评价:(1)完全缓解(complete response, CR):所有靶病灶中的肿瘤内动脉期增强消失;(2)部分缓解(partial remission, PR):靶病灶的大小或增强病变的所有直径之和至少减少30%;(3)疾病稳定(stable disease, SD):靶病灶的最小直径既没有达到PR,增大程度也未达到疾病进展的直径要求;(4)疾病进展(progressive disease, PD):肿瘤大小(靶病灶的所有增强病灶的直径之和以及病灶的最大径)增加至少20%,或出现新病灶。按随访结果将入组患者分为有效组34例(12例CR和22例PR),无效组30例(10例PD和20例SD)。

1.4 影像设备及检查方法

       本研究选用了3.0 T超导磁共振扫描仪(GE Signa Hdxt, USA),并采用32通道体部线圈。所有患者检查前进行呼吸训练,以配合检查中的闭气,避免由于呼吸伪影对肝脏病变显示产生影响。扫描序列及参数:(1)轴位T1WI,TR 375 ms,TE Minimum,FOV 40 cm× 40 cm,层厚5 mm,层间距1 mm,翻转角85°,NEX 0.5,扫描时长26 s;(2)轴位fat-saturated T2-weighted imaging,TR 4 000 ms,TE 80 ms,FOV 40 cm×40 cm,NEX 4,层厚5 mm,层间距1 mm,ETL 24,扫描时长1 min 37 s;(3)轴位DWI,TR 1675 ms,TE Minimum,FOV 40 cm× 40 cm,NEX 4,层厚5 mm,b值800 s/mm2,扫描时长28 s;(4)常规MRI平扫后行T1WI对比增强扫描,采用3D-LAVA序列,TR 4 s,TE 1.6 s,FOV 40 cm× 40 cm,使用高压注射器静脉注射Gd-DTPA对比剂(规格20 mL,广州康臣药业有限公司,中国),计量0.2 mmol/kg,注射对比剂后在12~16、53~57、120 s后行肝脏扫描,每期扫描时间为12~14 s,每期扫描84层,获得T1WI三期扫描图像(动脉期、静脉期及延迟期)。

1.5 图像分析及特征提取

       将所有患者图像从PACS上以BPM格式导出,保证所有图像的窗宽窗位一致,使用MaZda(version 4.6,http://www.eletel.p.lodz.pl/programy/mazda/)软件进行特征提取,在特征提取之前进行图像的灰度标准化并控制在(μ±3σ)水平,由两名分别具有5年和8年诊断经验的腹部影像诊断主治医师在每位患者TACE前后轴位T1WI增强扫描动脉期、门脉期及延迟期图像上选择病灶最大层面人工手动勾画ROI,首先由一名医师进行勾画,另一位高年资医师进行复核,如意见不统一,则共同讨论决定,在勾画ROI时不避开病灶内的任何成分,应包含整个病灶,包括囊变、钙化、血管等。软件自动计算出纹理参数,本研究分析提取病灶的一阶纹理特征,主要是直方图纹理参数,包括均值、方差、偏度、峰度以及第1、10、50、90和99百分位数值(分别记为Perc1、Perc10、Pere50、Perc90和Perc99),共9个纹理参数(图1)。

图1  女,66岁,乙肝肝硬化,肝细胞癌。经导管肝动脉化疗栓塞(TACE)术前行横轴位T1WI对比增强扫描后动脉期(1A)、静脉期(1B)、延迟期(1C)MRI图像,MaZda软件勾画感兴趣区示意图(1D),对应动脉期(1E)、静脉期(1F)、延迟期(1G)灰度直方图;TACE术后横轴位T1WI对比增强扫描动脉期(1H)、静脉期(1I)、延迟期(1J)MRI图像,MaZda软件勾画感兴趣区示意图(1K)。
Fig. 1  Female, 66 years old, hepatitis B, liver cirrhosis, hepatocellular carcinoma patient. Before transcatheter arterial chemoembolization (TACE), transverse T1WI contrast-enhanced scans are performed, followed by arterial phase (1A), venous phase (1B), and delayed phase (1C) MRI images, MaZda software is used to outline the region of interest (1D), corresponding grayscale histograms of arterial phase (1E), venous phase (1F), and delayed phase (1G); after TACE, transverse T1WI contrast-enhanced scans are performed, followed by arterial phase (1H), venous phase (1I), and delayed phase (1J) MRI images, MaZda software is used to outline the region of interest (1K).

1.6 统计学分析

       本研究所有数据均使用SPSS 26.0软件(SPSS, Chicago, IL, USA)和MedCalc 15.0软件进行统计分析。采用Kolmogorov-Smirnov检验测量数据是否符合正态分布,计量资料符合正态分布时使用均数±标准差(x¯±s)表示,不符合正态分布时采用中位数(上、下四分位数)表示。计数资料采用百分率表示。对两组间(有效组和无效组)的计数资料采用卡方检验,符合正态分布的两组间计量资料采用两独立样本t检验,不符合正态分布的计量资料则使用Mann-Whitney U检验。以P<0.05为差异有统计学意义。使用受试者工作特征(receiver operating characteristic, ROC)曲线分析差异有统计学意义的一阶纹理参数预测效能,使用DeLong检验评估曲线下面积(area under the curve, AUC)值差异。

2 结果

2.1 患者临床资料

       本研究最终纳入研究对象64例(其中,经病理证实35例,临床及影像学诊断29例),共64个病灶,男41例(共41个病灶),女23例(共23个病灶),年龄33~77(59.36±9.58)岁。

       依据mRECIST标准,将64例患者分为有效组(CR和PR)34例和无效组(PD和SD)30例。经统计学分析,两组间年龄、性别差异无统计学意义(P>0.05)(表1)。

表1  所有患者的基本特征
Tab. 1  Basic characteristics of all of the patients

2.2 TACE术前、后两组间各直方图特征参数对比

       结果显示,术前两组间动脉期均值、Perc50、Perc90、Perc99及静脉期Perc1、延迟期峰度值差异有统计学意义(P<0.05)。其中,无效组术前动脉期均值、Perc50、Perc90、Perc99及延迟期峰度显著高于有效组,静脉期Perc1低于有效组(P<0.05)。术后两组间延迟期均值、Perc50、Perc99差异有统计学意义(P<0.05),其余两组间TACE前后动脉期、静脉期及延迟期各直方图参数差异无统计学意义(P>0.05),其中,无效组术后延迟期均值、Perc50、Perc99显著高于有效组(P<0.05)(表2),有效组TACE术前、术后各直方图参数详见表3

表2  两组间各直方图参数比较
Tab. 2  Comparison of the various histogram feature parameters between the two groups
表3  有效组TACE治疗前、后各直方图参数比较
Tab. 3  Comparison of histogram feature parameters before and after TACE treatment in the effective group

2.3 肝癌TACE术后各直方图参数的早期疗效预测价值

       有效组和无效组间差异有统计学意义的各直方图参数进行ROC曲线分析显示,术前动脉期均值的预测效能最高,其AUC值为0.710(95% CI:0.583~0.817),其次分别为术后延迟期Perc99、术前动脉期Perc90、术后延迟期均值、术前延迟期峰度、术后延迟期Perc50、术前动脉期Perc50、术前动脉期Perc99、术前静脉期Perc1(图2表4)。

       术前动脉期均值、术前动脉期Perc50、术前动脉期Perc90、术前动脉期Perc99、术前静脉期Perc1、术前延迟期峰度的AUC分别为0.710、0.662、0.669、0.659、0.339、0.666,组间差异无统计学意义;术后延迟期均值、术后延迟期Perc50、术后延迟期Perc99的AUC分别为0.667、0.665、0.676,组间差异无统计学意义,详见表5

图2  各直方图参数受试者工作特征曲线图。TACE:经导管肝动脉化疗栓塞。
Fig. 2  Receiver operating characteristic curves of each histogram feature parameter. TACE: transcatheter arterial chemoembolization.
表4  各直方图参数的预测效能分析
Tab. 4  Results of the receiver operating characteristic curves for each histogram feature parameters
表5  TACE术前、术后差异有统计学意义的参数DeLong检验比较结果
Tab. 5  Comparison results of Delong test for parameters with statistically significant differences before and after TACE

3 讨论

       本研究以CE-MRI图像的一阶纹理参数分析为基础,将纳入的患者分为有效组和无效组,通过纹理分析软件对肝癌病灶动脉期、静脉期及延迟期的一阶纹理特征进行提取,共9个直方图特征参数,包括均值、方差、偏度、峰度以及Perc1、Perc10、Pere50、Perc90和Perc99,探讨各参数对肝癌TACE术后早期疗效的评估和预测价值。本研究首次将CE-MRI延迟期图像纹理参数引入,对比分析了两组间TACE前、后MRI动脉期、静脉期及延迟期图像的各参数之间的差异,并将术前及术后差异有统计学意义的参数进行ROC曲线分析,结果表明CE-MRI图像直方图参数对肝癌TACE术后疗效及预测评估有重要的意义,为临床治疗方案的选择提供参考。

3.1 CE-MRI直方图分析鉴别有效组和无效组临床意义

       纹理分析是近年来发展起来的一种新的图像分析技术,已经被广泛应用到各种肿瘤成像中,是从医学图像中提取定量参数来客观评估病变的特征。传统成像分析通常将肿瘤作为一个独立的整体,而纹理分析则不同,它将肿瘤分解成多个独立的像素,它通过量化图像中像素的灰阶变化,来评估肿瘤的异质性。均值、Perc1、Perc50、Perc90、Perc99这些纹理参数属于一阶纹理参数,反映了ROI内的灰度频度分布[24],已经广泛用于评估肿瘤的病理分级及治疗反应等[25, 26, 27]。本研究中,我们发现无效组术前动脉期均值、Perc50、Perc90、Perc99、延迟期峰度显著高于有效组,静脉期Perc1低于有效组(P<0.05)。均值反映的是图像ROI中像素的平均值,均值越大代表肿瘤的异质性越高;峰度反映的是频率分布的尖峰形状,峰度越大,提示变量分布在均值附近比较集中;n百分位数代表的是低于或等于n%的研究对象占的比例,病变的异质性越大,高百分位数差异越大,相反,低百分位数差异越小。本研究中无效组(PD+SD)均值、峰度显著大于有效组(CR+PR)。无效组包含了SD和PD,说明无效组肿瘤细胞的异型性较高,血供丰富,强化程度高,因而T1WI增强图像的像素平均值及峰度高于有效组,静脉期Perc1低于有效组(P<0.05)。我们还将TACE术后各直方图参数进行了比较,发现无效组术后延迟期均值、Perc50、Perc99显著高于有效组(P<0.05),我们推测术前无效组这些肿瘤异质性高的病灶在术后也有可能高于有效组,由于我们的研究病例数较少,后续我们将增加病例数进一步地探索和研究这一推测。另外,我们将有效组TACE术前、术后的直方图参数进行了对比,发现有效组TACE术前静脉期均值、Perc1、Perc10、Perc50、Perc99以及延迟期Perc1、Perc10高于术后(P<0.05),说明有效组在TACE治疗后肿瘤细胞的活性及肿瘤的异质性减低。

       有研究表明,在一定程度上,纹理特征值的大小能够反映肿瘤组织的异质性,肿瘤的异质性可能与肿瘤细胞浸润、异常血管形成、微血管形成和坏死有关[28, 29, 30],还有研究发现,与肿瘤异质性相关的纹理参数应该或多或少与肿瘤的组织学特征相关[31, 32, 33]。在肿瘤学中,对肿瘤组织的异质性评估是非常重要的,肿瘤的异质性程度是生存预后的决定因素[34, 35, 36]。在本研究中,上述TACE术前动脉期差异有统计学意义的一阶纹理参数值无效组高于有效组,因而,无效组的肿瘤异质性较大。无效组包含了肝癌TACE术后SD、PD,提示肿瘤内部的结构分布及微环境更为复杂,因此治疗效果不如有效组。这与以往的研究结果相似[9, 13, 28]。此外,肝癌95%~99%的血供来自肝动脉,这也是TACE治疗的理论基础,因此,笔者推测术前动脉期差异有统计学意义的纹理参数值更能客观地评估肝癌病灶的异质性。

3.2 CE-MRI直方图参数预测肝癌TACE术后早期疗效

       本研究将两组间差异有统计学意义的一阶纹理特征值进行ROC曲线分析,结果显示术前动脉期均值、Perc50、Perc9、Perc99和延迟期峰度以及术后延迟期均值、延迟期Perc50、延迟期Perc99对肝癌TACE术后早期疗效为无效组的预测有一定的价值,AUC均大于0.5,术前动脉期均值的AUC最大(0.710),说明术前动脉期均值对肝癌患者TACE术后早期疗效预测最佳,并能在TACE术前进行提前预测,为TACE治疗前提供有效的定量依据。YU等[37]通过基于CE-MRI图像的纹理分析预测TACE联合高强度聚焦超声治疗后早期疗效价值的研究发现,治疗前的部分纹理参数对术后非完全反应具有一定的价值,这与本研究得出的结果一致。因此在TACE实施前,准确评估治疗反应以预测疗效,可以为是否进行再次TACE治疗提供线索,进而指导后续治疗,以实现残余肿瘤的完全坏死[38]

3.3 本研究的局限性

       然而,本研究还存在以下几方面的局限性:首先,本研究中提取的直方图参数为肿瘤最大层面,获得的并不是肿瘤全域信号,难以全面地估肿瘤的异质性;其次,收集的病例数量较少,后续需扩大样本量继续深入研究。

4 结论

       综上所述,CE-MRI图像的一阶纹理参数能够对肝癌TACE术后早期肿瘤组织的特征进行早期评估及预测,不仅能够直观地观察肝癌TACE术后的影像学表现,还能通过纹理参数的变化来评估人眼难以分辨的肿瘤内部特征,对临床治疗方案的选择有重要的意义。

[1]
FORNER A, REIG M, BRUIX J. Hepatocellular carcinoma[J]. Lancet, 2018, 391(10127): 1301-1314. DOI: 10.1016/S0140-6736(18)30010-2.
[2]
NAN Y M, XU X Y, GAO Y H, et al. Consensus on the secondary prevention of primary liver cancer[J]. Hepatol Int, 2021, 15(6): 1289-1300. DOI: 10.1007/s12072-021-10259-7.
[3]
RAOUL J L, FORNER A, BOLONDI L, et al. Updated use of TACE for hepatocellular carcinoma treatment: how and when to use it based on clinical evidence[J]. Cancer Treat Rev, 2019, 72: 28-36. DOI: 10.1016/j.ctrv.2018.11.002.
[4]
ZHI W H, HOU J, FAN S P, et al. Delayed biopsy following completion of transarterial chemoembolization in patients with hepatocellular carcinoma: effects on pathological outcomes and its advantages[J]. J Cancer Res Ther, 2022, 18(5): 1346-1351. DOI: 10.4103/jcrt.jcrt_732_22.
[5]
BAERE T D, ARAI Y, LENCIONI R, et al. Treatment of liver tumors with lipiodol TACE: technical recommendations from experts opinion[J]. Cardiovasc Intervent Radiol, 2016, 39(3): 334-343. DOI: 10.1007/s00270-015-1208-y.
[6]
KIM K M, KIM J H, PARK I S, et al. Reappraisal of repeated transarterial chemoembolization in the treatment of hepatocellular carcinoma with portal vein invasion[J]. J Gastroenterol Hepatol, 2009, 24(5): 806-814. DOI: 10.1111/j.1440-1746.2008.05728.x.
[7]
LIU J, PEI Y G, ZHANG Y, et al. Predicting the prognosis of hepatocellular carcinoma with the treatment of transcatheter arterial chemoembolization combined with microwave ablation using pretreatment MR imaging texture features[J]. Abdom Radiol, 2021, 46(8): 3748-3757. DOI: 10.1007/s00261-020-02891-y.
[8]
LEWIS S, HECTORS S, TAOULI B. Radiomics of hepatocellular carcinoma[J]. Abdom Radiol (NY), 2021, 46(1): 111-123. DOI: 10.1007/s00261-019-02378-5.
[9]
MAO X N, GUO Y, WEN F, et al. Applying arterial enhancement fraction (AEF) texture features to predict the tumor response in hepatocellular carcinoma (HCC) treated with Transarterial chemoembolization (TACE)[J/OL]. Cancer Imaging, 2021, 21(1): 49 [2022-06-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8359085/. DOI: 10.1186/s40644-021-00418-2.
[10]
GU Y, HUANG H, TONG Q, et al. Multi-view radiomics feature fusion reveals distinct immuno-oncological characteristics and clinical prognoses in hepatocellular carcinoma[J/OL]. Cancers, 2023, 15(8): 2338. [2023-11-01]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10137067/. DOI: 10.3390/cancers15082338.
[11]
AKıNCı Ö, TÜRKOGLU F, NALBANT M O, et al. The effectiveness of volumetric MRI histogram analysis in renal cell carcinoma[J]. Acad Radiol, 2023, 30(Suppl 1): S278-S285. DOI: 10.1016/j.acra.2023.03.029.
[12]
CHEN S, HARMON S, PERK T, et al. Using neighborhood gray tone difference matrix texture features on dual time point PET/CT images to differentiate malignant from benign FDG-avid solitary pulmonary nodules[J/OL]. Cancer Imaging, 2019, 19(1): 56 [2022-06-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6697997/. DOI: 10.1186/s40644-019-0243-3.
[13]
GANESHAN B, SKOGEN K, PRESSNEY I, et al. Tumour heterogeneity in oesophageal cancer assessed by CT texture analysis: preliminary evidence of an association with tumour metabolism, stage, and survival[J]. Clin Radiol, 2012, 67(2): 157-164. DOI: 10.1016/j.crad.2011.08.012.
[14]
陈瑾, 王海屹, 叶慧义. 纹理分析在肿瘤影像学中的研究进展[J]. 中华放射学杂志, 2017, 51(12): 979-982. DOI: 10.3760/cma.j.issn.1005-1201.2017.12.020.
CHEN J, WANG H Y, YE H Y. Research progress of texture analysis in tumor imaging[J]. Chin J Radiol, 2017, 51(12): 979-982. DOI: 10.3760/cma.j.issn.1005-1201.2017.12.020.
[15]
SHAO C C, ZHAO F, YU Y F, et al. Value of perfusion parameters and histogram analysis of triphasic computed tomography in pre-operative prediction of histological grade of hepatocellular carcinoma[J]. Chin Med J, 2021, 134(10): 1181-1190. DOI: 10.1097/CM9.0000000000001446.
[16]
ZHOU W, LV Y Z, HU X M, et al. Study on the changes of CT texture parameters before and after HCC treatment in the efficacy evaluation and survival predication of patients with HCC[J/OL]. Front Oncol, 2022, 12: 957737 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9650244/. DOI: 10.3389/fonc.2022.957737.
[17]
LI L X, YANG Z H, ZHENG Y F, et al. Identification of an endoplasmic reticulum stress-related signature associated with clinical prognosis and immune therapy in glioma[J/OL]. BMC Neurol, 2022, 22(1): 192 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9131635/. DOI: 10.1186/s12883-022-02709-y.
[18]
CHOI J M, YU J S, CHO E S, et al. Texture analysis of hepatocellular carcinoma on magnetic resonance imaging: assessment for performance in predicting histopathologic grade[J]. J Comput Assist Tomogr, 2020, 44(6): 901-910. DOI: 10.1097/RCT.0000000000001087.
[19]
WANG Y M, PAN X P, LIN H, et al. Multi-scale pathology image texture signature is a prognostic factor for resectable lung adenocarcinoma: a multi-center, retrospective study[J/OL]. J Transl Med, 2022, 20(1): 595 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9749333/. DOI: 10.1186/s12967-022-03777-x.
[20]
PARK H J, KIM J H, CHOI S Y, et al. Prediction of therapeutic response of hepatocellular carcinoma to transcatheter arterial chemoembolization based on pretherapeutic dynamic CT and textural findings[J]. AJR Am J Roentgenol, 2017, 209(4): W211-W220. DOI: 10.2214/AJR.16.17398.
[21]
东贤进, 高军林, 苏伟, 等. 基于术前MRI的纹理分析对47例肝细胞癌经动脉化疗栓塞术治疗后早期复发的预测分析[J]. 肿瘤学杂志, 2023, 29(2): 140-144. DOI: 10.11735/j.issn.1671-170X.2023.02.B009.
DONG X J, GAO J L, SU W, et al. Texture analysis based on preoperative MRI to predict 47 cases of liver cancer early recurrence after transarterial chemoem-bolization[J]. J Chin Oncol, 2023, 29(2): 140-144. DOI: 10.11735/j.issn.1671-170X.2023.02.B009.
[22]
《原发性肝癌诊疗规范(年版)》编写专家委员会. 原发性肝癌诊疗规范(2019年版)[J]. 中国临床医学, 2020, 27(1): 140-156. DOI: 10.12025/j.issn.1008-6358.2020.20200065.
Expert Committee for the Compilation of the "Diagnosis and Treatment Standards for Primary Liver Cancer (2019 Edition)". Diagnostic and therapeutic criteria for primary liver cancer (2019 edition)[J]. Chin J Clin Med, 2020, 27(1): 140-156. DOI: 10.12025/j.issn.1008-6358.2020.20200065.
[23]
LENCIONI R, LLOVET J M. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma[J]. Semin Liver Dis, 2010, 30(1): 52-60. DOI: 10.1055/s-0030-1247132.
[24]
VOSSHENRICH J, ZECH C J, HEYE T, et al. Response prediction of hepatocellular carcinoma undergoing transcatheter arterial chemoembolization: unlocking the potential of CT texture analysis through nested decision tree models[J]. Eur Radiol, 2021, 31(6): 4367-4376. DOI: 10.1007/s00330-020-07511-3.
[25]
许岂豪, 赵莹, 王悦, 等. R2*图纹理分析预测肝细胞癌肝切除术后早期复发的价值[J]. 磁共振成像, 2022, 13(12): 87-92. DOI: 10.12015/issn.1674-8034.2022.12.015.
XU Q H, ZHAO Y, WANG Y, et al. Value of texture analysis based on R2*map for predicting early recurrence of HCC after hepatectomy[J]. Chin J Magn Reson Imag, 2022, 13(12): 87-92. DOI: 10.12015/issn.1674-8034.2022.12.015.
[26]
BURCHARDT E, BOS-LIEDKE A, SERKOWSKA K, et al. Value of[18F]FDG PET/CT radiomic parameters in the context of response to chemotherapy in advanced cervical cancer[J/OL]. Sci Rep, 2023, 13(1): 9092 [2023-10-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10241798/. DOI: 10.1038/s41598-023-35843-9.
[27]
CHOI B K, PARK S, LEE G, et al. Can CT texture analysis parameters be used as imaging biomarkers for prediction of malignancy in canine splenic tumors?[J]. Vet Radiol Ultrasound, 2023, 64(2): 224-232. DOI: 10.1111/vru.13175.
[28]
LI M, FU S R, ZHU Y J, et al. Computed tomography texture analysis to facilitate therapeutic decision making in hepatocellular carcinoma[J]. Oncotarget, 2016, 7(11): 13248-13259. DOI: 10.18632/oncotarget.7467.
[29]
MĂRGINEAN L, ȘTEFAN P A, LEBOVICI A, et al. CT in the differentiation of gliomas from brain metastases: the radiomics analysis of the peritumoral zone[J/OL]. Brain Sci, 2022, 12(1): 109 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8774238/. DOI: 10.3390/brainsci12010109.
[30]
LUO Y, QIN L K, YAN J W, et al. Classification of contrast-enhanced ultrasonograms in rectal cancer according to tumor inhomogeneity using machine learning-based texture analysis[J]. Transl Cancer Res, 2022, 11(5): 1053-1063. DOI: 10.21037/tcr-21-2362.
[31]
RATHORE S, NIAZI T, IFTIKHAR M A, et al. Glioma grading via analysis of digital pathology images using machine learning[J/OL]. Cancers, 2020, 12(3): 578 [2022-06-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7139732/. DOI: 10.3390/cancers12030578.
[32]
YAŞAR S, VOYVODA N, VOYVODA B, et al. Using texture analysis as a predictive factor of subtype, grade and stage of renal cell carcinoma[J]. Abdom Radiol, 2020, 45(11): 3821-3830. DOI: 10.1007/s00261-020-02495-6.
[33]
LEO A D, VARA G, PACCAPELO A, et al. Computerized tomography texture analysis of pheochromocytoma: relationship with hormonal and histopathological data[J]. J Endocrinol Invest, 2022, 45(10): 1935-1944. DOI: 10.1007/s40618-022-01826-2.
[34]
GRANATA V, FUSCO R, SETOLA S V, et al. An update on radiomics techniques in primary liver cancers[J/OL]. Infect Agent Cancer, 2022, 17(1): 6 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8897888/. DOI: 10.1186/s13027-022-00422-6.
[35]
DAVNALL F, YIP C S, LJUNGQVIST G, et al. Assessment of tumor heterogeneity: an emerging imaging tool for clinical practice?[J]. Insights Imaging, 2012, 3(6): 573-589. DOI: 10.1007/s13244-012-0196-6.
[36]
MALSHY K, AMIEL G E, HERSHKOVITZ D, et al. Association between nuclear morphometry parameters and gleason grade in patients with prostatic cancer[J/OL]. Diagnostics, 2022, 12(6): 1356 [2023-01-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9222000/. DOI: 10.3390/diagnostics12061356.
[37]
YU J Y, ZHANG H P, TANG Z Y, et al. Value of texture analysis based on enhanced MRI for predicting an early therapeutic response to transcatheter arterial chemoembolisation combined with high-intensity focused ultrasound treatment in hepatocellular carcinoma[J/OL]. Clin Radiol, 2018, 73(8): 758.e9-758.e18 [2022-06-30]. https://sci-hub.se/10.1016/j.crad.2018.04.013. DOI: 10.1016/j.crad.2018.04.013.
[38]
TIPALDI M A, RONCONI E, LUCERTINI E, et al. Hepatocellular carcinoma drug-eluting bead transarterial chemoembolization (DEB-TACE): outcome analysis using a model based on pre-treatment CT texture features[J/OL]. Diagnostics, 2021, 11(6): 956 [2022-06-30]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8226518/. DOI: 10.3390/diagnostics11060956.

上一篇 心脏磁共振对房颤患者左心房/左心耳血栓诊断价值的系统评价与Meta分析
下一篇 基于ZOOM-mDixon的T2*/R2*成像术前预测胰腺导管腺癌淋巴结转移的价值
  
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