CT术前预测肝细胞癌微血管侵犯的研究进展
Advances in CT Preoperative Prediction of Microvascular Invasion in Hepatocellular Carcinoma
DOI: 10.12677/ACM.2023.1351040, PDF, HTML, XML, 下载: 221  浏览: 346 
作者: 贺小娟, 郭大静*:重庆医科大学附属第二医院放射科,重庆
关键词: 肝细胞癌微血管侵犯计算机断层扫描术前诊断Hepatocellular Carcinoma Microvascular Invasion Computer Tomography Preoperative Diagnosis
摘要: 肝细胞癌(HCC)是世界上最常见的恶性肿瘤之一,预后差、病死率高。微血管侵犯(MVI)是HCC患者无病生存和总体生存的独立因素。因此,术前有效预测MVI状态将可以帮助临床医生制定治疗方案,尤其是手术方式和术后联合治疗的选择,由此来改善患者的预后。本文将对目前基于CT检查术前预测肝细胞癌患者MVI的研究进展进行综述。
Abstract: Hepatocellular carcinoma (HCC) is one of the most common malignant tumors in the world, with poor prognosis and high mortality. Microvascular invasion (MVI) is an independent factor for dis-ease free survival and overall survival in HCC patients. Therefore, effective preoperative prediction of MVI status will help clinicians formulate treatment plans, especially surgical methods and post-operative combination therapy options, thereby improving the prognosis of patients. This review summarizes the current research progress in preoperative diagnosis of MVI in HCC patients based on computed tomography (CT) examination.
文章引用:贺小娟, 郭大静. CT术前预测肝细胞癌微血管侵犯的研究进展[J]. 临床医学进展, 2023, 13(5): 7454-7460. https://doi.org/10.12677/ACM.2023.1351040

1. 引言

原发性肝癌(Primary hepatic carcinoma, PHC)是最常见的恶性肿瘤之一,其全球发病率和死亡率中分别排名第六和第四,尤其是在亚洲地区,其发病率远超过其他地区。原发性肝癌包括肝细胞癌(HCC)、肝内胆管细胞癌(ICC)、混合型(HCC-ICC),以及其他罕见的亚型,HCC是其最常见的亚型,约占PHC的80% [1] [2] 。HCC的主要危险因素是慢性乙型肝炎病毒(HBV)或丙型肝炎病毒(HCV)感染、黄曲霉毒素污染的食物、大量饮酒、超重、2型糖尿病和吸烟。在我国,其主要危险因素是慢性HBV感染。HCC起病隐匿,确诊时多属中晚期,因而其治疗效果和预后较差,病死率高。目前,已有多种方法应用于HCC的治疗,其主要治疗方法包括有手术切除、肝移植、射频消融、经导管动脉化疗栓塞(TACE)、靶向治疗和免疫治疗 [3] 。其中,手术切除仍是HCC最重要的治疗方式,但手术后复发率仍较高,其5年复发率达70% [4] 。影响患者复发的因素很多,有大量研究表明,微血管侵犯(Microvascular invasion, MVI)是HCC患者术后复发的危险因素之一。术前预测HCC的MVI状态是近年来肝癌领域的研究热点,通过术前的预测结果将有利于临床手术方案的建立,以及后续综合治疗的选择,从而改善患者的预后。影像学检查在HCC的术前诊断中起着至关重要的作用。计算机断层扫描(Computed tomography, CT)应用广泛、相对经济,它可以对初检的肝脏不确定结节做出初步的定性诊断,美国肝病研究组织(American Association for the Study of Liver Diseases, AASLD)在指南中指出在肝硬化和疑似HCC患者中,诊断性成像用于非侵入性验证HCC的存在并确定其程度 [3] ,美国放射学会也制定了肝脏影像报告及数据系统(Liver imaging report and data system, LI-RADS)用于HCC的影像学诊断,术后定期的CT复查也可帮助临床了解预后情况,进行相应的治疗选择,由此可见CT在HCC管理各个阶段发挥着重要作用。因此,本综述总结了目前基于CT检查术前诊断HCC患者MVI的研究进展。

2. MVI的定义

血管侵犯(Vascular invasion)通常分为两种,一种是肉眼可见,即在大体检查时或者影像学检查时可见;另一种是显微镜下,即仅在显微镜下可见。即MVI主要是指在显微镜下于内细胞衬覆的血管腔内见到癌细胞巢团。多见于癌旁肝组织内的门静脉小分支(含肿瘤包膜内血管),肝静脉分支作为肝癌次要的出瘤血管也可发生MVI,偶可见到肝癌侵犯肝动脉、胆管以及淋巴管等脉管小分支。我国的病理学家和临床专家在原发性肝癌规范化病理诊断指南 [5] 中建议所有肝癌标本均应基于7点基线样本采集方案进行采样,以对原发性肝癌进行更准确的病理诊断。并且指出MVI应在所有组织切片中进行评估,MVI分级系统分为三个级别:M0:未发现MVI;M1 (低危组):≤5个MVI,且发生于近癌旁肝组织区域(≤1 cm);M2 (高危组):>5个MVI,或MVI发生于远癌旁肝组织区域(>1 cm)。当癌旁肝组织内的卫星灶与MVI难以区分时,可一并计入MVI分级。除此之外,还有其他学者对于MVI分型提出了不同的分类模式 [6] [7] [8] 。

3. MVI的临床意义

文献资料显示,HCC患者MVI的发生率为15.0%~57.1% [9] [10] 。MVI的存在是HCC侵袭性行为的病理特征,预示着肝切除和移植后具有较差的预后。Xu等人研究中的多因素Cox回归分析结果显示,MVI的HR值为1.686,说明MVI是HCC患者肝切除术后晚期复发的独立危险因素 [11] ,在监测MVI阳性患者时,对其进行相应的干预治疗将可以改善患者的预后。在肝脏双重血供的生理基础下,富血供的HCC更易发生MVI,尤其是富含高侵袭性细胞的瘤周区域。一些研究提出,对于MVI阳性的HCC患者,扩大手术切除的切缘将改善患者术后的治疗结果。Han等人研究了MVI阴性和阳性患者在进行窄切缘和宽切缘肝切除术后的预后情况,结果表明,宽切缘术后患者的总生存期和无复发生存期均长于窄切缘的患者,并且HCC患者合并MVI阳性和窄切缘切除两种因素可使术后死亡和复发的风险增加约2倍 [12] 。同时,MVI的存在状态对于患者的治疗方式的选择也具有一定的临床意义,一些研究表明,术后TACE等辅助治疗以及靶向、免疫治疗等综合治疗可以改善合并MVI患者的生存结果 [13] [14] 。

4. 基于CT的MVI的术前诊断

4.1. 影像学特征

通过CT检查的图像可以获取患者肿瘤的大小、肿瘤数目和肿瘤形态,并且还能进一步得到肿瘤边缘和包膜的情况、瘤周组织强化方式及肿瘤内动脉等影像学特征,从而帮助术前评估MVI的存在状态。如Song等人 [15] 的研究中表明影像学观察肿瘤边缘不光滑、包膜不完整对HCC术前MVI的预测有重要意义。Banerjee等人 [16] 则提出了一种非侵入性的基于对比增强CT影像学特征联合基因学的生物标志物,称为放射基因组静脉侵袭(Radiogenomic venous invasion, RVI)。RVI由三个独立的成像特征组成:“内动脉(Internal arteries)”是在成像的静脉期肿瘤内持续存在离散的动脉增强;“低密度晕(hypodense Halo)”是部分或全部包围肿瘤的低密度边缘;“肿瘤–肝脏差异(tumor-liver difference)”是指肿瘤与相邻肝实质之间在没有低密度晕的情况下,呈局灶性或周向性衰减的急剧转变。RVI阳性被定义为存在内动脉、不存在低密度晕和肿瘤–肝脏差异。研究结果显示RVI预测MVI的准确性、敏感性和特异性达到了89%、76%和84%。基于成像特征和基因表达之间关联的静脉侵犯的双预测因子(two-trait predictor of venous invasion, TTPVI)也在研究中被证明是可靠的预测指标 [17] 。在单因素分析中,TTPVI与MVI呈统计学显著相关,在多因素逻辑回归分析中其OR值为4.802,表明TTPVI是MVI的重要独立预测因素。

4.2. 影像组学

除了常规可读取的影像信息,影像图像中还具有能反映肿瘤细胞生物学行为等的潜在信息。影像组学则是通过高通量计算机算法,从传统的二维或三维的影像图像中(超声、CT、MRI、PET/CT等)自动提取相应的定量影像组学特征用于统计分析,并且进一步构建预测模型 [18] [19] [20] 。这样将能提高图像信息的利用率,也能提高影像诊断的准确性。目前,影像组学方法已经应用于HCC的诊断、鉴别诊断、病理分级及预后评估等多个方面 [21] [22] [23] [24] [25] 。影像组学研究的主要工作流程为:1) 图像采集及预处理;2) 图像分割:通过自动分割方法或半自动分割方法或由放射诊断医师手动勾画感兴趣区;3) 组学特征提取及选择:通常包括形态学等相关描述符、纹理特征、直方图特征等;4) 统计分析、建立模型:经过一系列降维选择的组学特征可通过不同机器学习(Machine Learning, ML)算法建立模型。

Xu等人 [26] 根据7260个组学特征的选择设置了MVI相关的放射组学评分(R-scores)、整合15个临床因素和12个放射学评分建立了一个放射学–放射组学(radiographic-radiomics, RR)预测模型,该模型在训练/验证集中的AUC值为0.909,在测试集中为0.889,可以识别88%以上的MVI阳性病例,特异性为76.8%~79.2%。研究指出在7000多个放射学特征中,与肿瘤大小(如形状表面积)和异质性(如灰度级不均匀性、小波繁忙性、复杂性和熵)相关的因素是预测组织学MVI的最重要组成部分。Zhang等人 [27] 基于CT图像的放射组学特征和临床因素构建了两个分类器,一个用于术前评估HCC患者的MVI状态,一个用于进一步对MVI风险进行分层,最终用于预测MVI状态的分类器的AUC值达到了0.796以上,用于风险分层的分类器的AUC值达到了0.740以上。因为MVI通常发生在肿瘤/非肿瘤界面附近的区域,关于瘤周组织的研究也是热点之一。Zhang等人则利用了对比增强CT的不同期相与不同瘤周边缘距离组合来构建模型用于MVI的预测。结果显示,从门静脉期提取的肿瘤核心和瘤周12 mm边距的图像特征显示出主要的预测性能,其提取的特征包括一个基于灰度级大小区域矩阵(GLSZM)的特征和四个基于一阶的特征。融合模型在独立测试集上获得了最佳的预测性能,AUC值为0.81,准确率为78.3%,灵敏度为81.8%,特异性为75%。影像组学方法多样,尚缺乏统一的标准,大多数影像组学研究依赖人工手动分割病灶,操作上耗时耗力,并且影像组学提取的定量图像特征也有限,提取的特征也可能与图像本身存在视觉感知的差异,目前影像组学的应用范围也主要在于进行病灶的分类诊断。

4.3. 深度学习

随着人工智能(Artificial Intelligence, AI)的发展,深度学习(Deep learning, DL)技术逐渐应用于医学领域 [28] [29] 。DL是ML的一个新领域,也是近年AI相关研究的热点。DL源于对人工神经网络的研究,它使用了多层次的非线性信息处理和抽象,用于实现有监督或无监督的特征学习等任务。它通过建立类似于人脑的分层的分析学习模式,以模拟人脑自动学习、获取、分析、处理和解释数据,将“低层”特征组合转化为抽象的“高层”的特征,从而完成更复杂的任务。相较于传统的ML算法,DL不再需要手动选择特征和分类器,而是通过算法自动学习来提取特征,这使得DL在处理复杂数据集或大数据集上更具有优势 [30] 。目前在医学影像学领域广泛应用的有卷积神经网络(Convolution neural network, CNN),生成式对抗网络(Generative adversarial nets, GAN),循环神经网络模型(Recurrent neural nets, RNN)等。

近几年已有学者将DL运用到HCC术前诊断及预后评估的研究中 [31] - [36] 。Jiang等人 [37] 建立了基于放射组学特征、放射学特征和临床变量的XGBoost模型以及一个三维卷积神经网络(3D-CNN)来预测MVI状态并对者两种模型的性能进行比较。结果显示,3D-CNN的模型性能稍优于XGBoost模型的性能,在训练集和验证集中的AUC值达到了0.9以上。Liu [38] 等人也开发了一个基于DL的框架,通过使用动脉期(AP)的CT图像进行术前MVI预测,图像仅作简单的肿瘤标记,无需手动特征提取。研究利用AP图像和/或患者临床因素(CF)构建了深度学习(ResNet-18)和机器学习(支持向量机)模型,并对它们的性能进行了系统比较。随后,使用梯度加权类激活图(Grad-CAM)对最佳模型的可解释性进行可视化。结果表明,利用AP图像和患者临床因素建立的ResNet-18模型优于其他模型,其AUC值最高为0.845,Grad-CAM模型解释显示ResNet-18模型捕捉并学习了CT图像上与MVI相关的成像特征。综上,说明利用DL在术前诊断MVI方面是可行的,并且相较于传统的ML,在进一步提高预测性能上DL还具有更大的潜能。

随着研究进一步的展开,DL不仅可以用于分类预测,在整个研究流程中,我们还可以将DL应用到图像分割过程 [39] [40] [41] [42] 。Wang等人 [43] 提出了一种新的基于CT的MVI-Mind的端到端深度学习策略,该策略集成了数据预处理、病变和其他区域的自动分割、自动特征提取和MVI预测。分割和预测模块分别采用了轻量级变换器和卷积神经网络(CNN)。测试结果表明,MVI-Mind在分割和预测方面都具有了良好的性能。分割模块的平均并集交集(mIoU)为0.9006,预测模块的AUC值达到了0.9223。并且其端到端的预测速度可达到45.8~62.4 s/人。由此可见,随着DL的应用,不仅可以提高基于影像学检查对分类预测和预后评估的效能,还可以进一步扩大其应用功能,逐渐可以实现从图像采集、数据处理到结果输出全流程的自动化。

5. 总结与展望

综上所述,CT术前预测肝细胞肝癌MVI仍具有临床价值。传统通过影像学特征进行术前诊断,具有直观性和可操作性,但对放射诊断医师的经验要求较高。近年来人工智能在医学上的广泛应用,使得放射组学的研究已有显著的成果,随着深度学习技术的挖掘和探索,将来更加全面的多阶段多功能的综合模型将可能实现,从而达到在术前能快速且有效地预测HCC相关指标,并将其广泛运用到实际临床工作中。但目前,对其标准化、稳定性和泛化能力还有待进一步的研究。

NOTES

*通讯作者。

参考文献

[1] Sung, H., Ferlay, J., Siegel, R.L., Laversanne, M., Soerjomataram, I., Jemal, A. and Bray, F. (2021) Global Cancer Sta-tistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clinicians, 71, 209-249.
https://doi.org/10.3322/caac.21660
[2] 中华人民共和国国家卫生健康委员会医政医管局. 原发性肝癌诊疗规范(2019年版) [J]. 临床肝胆病杂志, 2020, 36(2): 277-292.
[3] Heimbach, J.K., Kulik, L.M., Finn, R.S., et al. (2018) AASLD Guidelines for the Treatment of Hepato-cellular Carcinoma. Hepatology (Baltimore, Md.), 67, 358-380.
https://doi.org/10.1002/hep.29086
[4] Lim, K.C., Chow, P.K., Allen, J.C., Siddiqui, F.J., Chan, E.S. and Tan, S.B. (2012) Systematic Review of Outcomes of Liver Re-section for Early Hepatocellular Carcinoma within the Milan Criteria. The British Journal of Surgery, 99, 1622-1629.
https://doi.org/10.1002/bjs.8915
[5] 吴孟超, 汤钊猷, 刘彤华, 等. 原发性肝癌规范化病理诊断指南(2015年版) [J]. 临床肝胆病杂志, 2015, 31(6): 833-839.
[6] Feng, L.H., Dong, H., Lau, W.Y., et al. (2017) Novel Micro-vascular Invasion-Based Prognostic Nomograms to Predict Survival Outcomes in Patients after R0 Resection for Hepa-tocellular Carcinoma. Journal of Cancer Research and Clinical Oncology, 143, 293-303.
https://doi.org/10.1007/s00432-016-2286-1
[7] Roayaie, S., Blume, I.N., Thung, S.N., et al. (2009) A System of Classifying Microvascular Invasion to Predict Outcome after Resection in Patients with Hepatocellular Carcinoma. Gas-troenterology, 137, 850-855.
https://doi.org/10.1053/j.gastro.2009.06.003
[8] Zhao, H., Chen, C., Fu, X., Yan, X., Jia, W., Mao, L., Jin, H. and Qiu, Y. (2017) Prognostic Value of a Novel Risk Classification of Microvascular Invasion in Patients with Hepatocellu-lar Carcinoma after Resection. Oncotarget, 8, 5474-5486.
https://doi.org/10.18632/oncotarget.12547
[9] Lei, Z., Li, J., Wu, D., Xia, Y., Wang, Q., Si, A., Wang, K., et al. (2016) Nomogram for Preoperative Estimation of Microvascular Invasion Risk in Hepatitis B Virus-Related Hepatocel-lular Carcinoma within the Milan Criteria. JAMA Surgery, 151, 356-363.
https://doi.org/10.1001/jamasurg.2015.4257
[10] Rodríguez-Perálvarez, M., Luong, T.V., Andreana, L., Meyer, T., Dhillon, A.P. and Burroughs, A.K. (2013) A Systematic Review of Microvascular Invasion in Hepatocellular Carcinoma: Diagnostic and Prognostic Variability. Annals of Surgical Oncology, 20, 325-339.
https://doi.org/10.1245/s10434-012-2513-1
[11] Xu, X.F., Xing, H., Han, J., Li, Z.L., Lau, W.Y., Zhou, Y.H., Gu, W.M., et al. (2019) Risk Factors, Patterns, and Outcomes of Late Recurrence after Liver Resection for Hepatocellular Carcinoma: A Multicenter Study from China. JAMA Surgery, 154, 209-217.
https://doi.org/10.1001/jamasurg.2018.4334
[12] Han, J., Li, Z.L., Xing, H., Wu, H., Zhu, P., Lau, W.Y., et al. (2019) The Impact of Resection Margin and Microvascular Invasion on Long-Term Prognosis after Curative Resection of Hepatocellular Carcinoma: A Multi-Institutional Study. HPB: The Official Journal of the International Hepato Pancreato Biliary Association, 21, 962-971.
https://doi.org/10.1016/j.hpb.2018.11.005
[13] Wang, H., Du, P.C., Wu, M.C. and Cong, W.M. (2018) Postopera-tive Adjuvant Transarterial Chemoembolization for Multinodular Hepatocellular Carcinoma within the Barcelona Clinic Liver Cancer Early Stage and Microvascular Invasion. Hepatobiliary Surgery and Nutrition, 7, 418-428.
https://doi.org/10.21037/hbsn.2018.09.05
[14] Wang, L., Chen, B., Li, Z., Yao, X., Liu, M., Rong, W., Wu, F., et al. (2019) Optimal Postoperative Adjuvant Treatment Strategy for HBV-Related Hepatocellular Carcinoma with Micro-vascular Invasion: A Propensity Score Analysis. OncoTargets and Therapy, 12, 1237-1247.
https://doi.org/10.2147/OTT.S179247
[15] Song, L., Li, J.W. and Luo, Y. (2021) The Importance of a Nonsmooth Tumor Margin and Incomplete Tumor Capsule in Predicting HCC Microvascular Invasion on Preoperative Imaging Ex-amination: A Systematic Review and Meta-Analysis. Clinical Imaging, 76, 77-82.
https://doi.org/10.1016/j.clinimag.2020.11.057
[16] Banerjee, S., Wang, D.S., Kim, H.J., Sirlin, C.B., Chan, M.G., Korn, R.L., Rutman, A.M., et al. (2015) A Computed Tomography Radiogenomic Biomarker Predicts Microvascular In-vasion and Clinical Outcomes in Hepatocellular Carcinoma. Hepatology (Baltimore, Md.), 62, 792-800.
https://doi.org/10.1002/hep.27877
[17] Zhang, T., Pandey, G., Xu, L., Chen, W., Gu, L., Wu, Y. and Chen, X. (2020) The Value of TTPVI in Prediction of Microvascular Invasion in Hepatocellular Carcinoma. Cancer Management and Research, 12, 4097-4105.
https://doi.org/10.2147/CMAR.S245475
[18] Lambin, P., Rios-Velazquez, E., Leijenaar, R., Carvalho, S., van Stiphout, R.G., Granton, P., Zegers, C.M., et al. (2012) Radiomics: Extracting More Information from Medical Images Using Advanced Feature Analysis. European Journal of Cancer (Oxford, England: 1990), 48, 441-446.
https://doi.org/10.1016/j.ejca.2011.11.036
[19] Gillies, R., Kinahan, P. and Hricak, H. (2016) Radiomics: Images Are More than Pictures, They Are Data. Radiology, 278, 563-577.
https://doi.org/10.1148/radiol.2015151169
[20] Lambin, P., Leijenaar, R.T.H., Deist, T.M., Peerlings, J., de Jong, E.E.C., van Timmeren, J., Sanduleanu, S., et al. (2017) Radiomics: The Bridge between Medical Imaging and Personal-ized Medicine. Nature Reviews. Clinical Oncology, 14, 749-762.
https://doi.org/10.1038/nrclinonc.2017.141
[21] Ji, G.W., Zhu, F.P., Xu, Q., Wang, K., Wu, M.Y., Tang, W.W., Li, X.C. and Wang, X.H. (2019) Machine-Learning Analy-sis of Contrast-Enhanced CT Radiomics Predicts Recurrence of Hepatocellular Carcinoma after Resection: A Mul-ti-Institutional Study. EBioMedicine, 50, 156-165.
https://doi.org/10.1016/j.ebiom.2019.10.057
[22] Shan, Q.Y., Hu, H.T., Feng, S.T., Peng, Z.P., Chen, S.L., Zhou, Q., Li, X., et al. (2019) CT-Based Peritumoral Radiomics Signatures to Predict Early Recurrence in Hepatocellular Carcinoma after Curative Tumor Resection or Ablation. Cancer Imaging, 19, Article No. 11.
https://doi.org/10.1186/s40644-019-0197-5
[23] Mokrane, F.Z., Lu, L., Vavasseur, A., Otal, P., Peron, J.M., Luk, L., et al. (2020) Radiomics Machine-Learning Signature for Diagnosis of Hepatocellular Carcinoma in Cirrhotic Patients with Indeterminate Liver Nodules. European Radiology, 30, 558-570.
https://doi.org/10.1007/s00330-019-06347-w
[24] Chong, H.H., Yang, L., Sheng, R.F., Yu, Y.L., Wu, D.J., Rao, S.X., Yang, C. and Zeng, M.S. (2021) Multi-Scale and Multi-Parametric Radiomics of Gadoxetate Disodium-Enhanced MRI Predicts Microvascular Invasion and Outcome in Patients with Solitary Hepatocellular Carcinoma ≤ 5 cm. European Radiology, 31, 4824-4838.
https://doi.org/10.1007/s00330-020-07601-2
[25] Hu, M.J., Yu, Y.X., Fan, Y.F. and Hu, C.H. (2021) CT-Based Radiomics Model to Distinguish Necrotic Hepatocellular Carcinoma from Pyogenic Liver Abscess. Clinical Radiology, 76, 161.e11-161.e17.
https://doi.org/10.1016/j.crad.2020.11.002
[26] Xu, X., Zhang, H.L., Liu, Q.P., Sun, S.W., Zhang, J., Zhu, F.P., Yang, G., et al. (2019) Radiomic Analysis of Contrast-Enhanced CT Predicts Microvascular Invasion and Outcome in Hepatocellular Carcinoma. Journal of Hepatology, 70, 1133-1144.
https://doi.org/10.1016/j.jhep.2019.02.023
[27] Zhang, X., Ruan, S., Xiao, W., Shao, J., Tian, W., Liu, W., Zhang, Z., et al. (2020) Contrast-Enhanced CT Radiomics for Preoperative Evaluation of Microvascular Invasion in Hepatocel-lular Carcinoma: A Two-Center Study. Clinical and Translational Medicine, 10, e111.
https://doi.org/10.1002/ctm2.111
[28] Greenspan, H., Ginneken, B.V. and Summers, R.M. (2016) Guest Editorial Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique. IEEE Transactions on Medical Imaging, 35, 1153-1159.
https://doi.org/10.1109/TMI.2016.2553401
[29] Huang, S., Yang, J., Fong, S. and Zhao, Q. (2020) Artificial Intel-ligence in Cancer Diagnosis and Prognosis: Opportunities and Challenges. Cancer Letters, 471, 61-71.
https://doi.org/10.1016/j.canlet.2019.12.007
[30] Zhu, Y., Wang, M., Yin, X., Zhang, J., Meijering, E. and Hu, J. (2022) Deep Learning in Diverse Intelligent Sensor Based Systems. Sensors (Basel, Switzerland), 23, Article No. 62.
https://doi.org/10.3390/s23010062
[31] Ioannou, G.N., Tang, W., Beste, L.A., Tincopa, M.A., Su, G.L., Van, T., Tapper, E.B., et al. (2020) Assessment of a Deep Learning Model to Predict Hepatocellular Carcinoma in Patients with Hepatitis C Cirrhosis. JAMA Network Open, 3, e2015626.
https://doi.org/10.1001/jamanetworkopen.2020.15626
[32] Wu, Y., White, G.M., Cornelius, T., Gowdar, I., Ansari, M.H., Supanich, M.P. and Deng, J. (2020) Deep Learning LI-RADS Grading System Based on Contrast Enhanced Mul-tiphase MRI for Differentiation between LR-3 and LR-4/LR-5 Liver Tumors. Annals of Translational Medicine, 8, 701.
https://doi.org/10.21037/atm.2019.12.151
[33] Liu, Q.P., Xu, X., Zhu, F.P., Zhang, Y.D. and Liu, X.S. (2020) Prediction of Prognostic Risk Factors in Hepatocellular Carcinoma with Transarterial Chemoembolization Using Mul-ti-Modal Multi-Task Deep Learning. EClinicalMedicine, 23, Article ID: 100379.
https://doi.org/10.1016/j.eclinm.2020.100379
[34] Wang, W., Chen, Q., Iwamoto, Y., Han, X., Zhang, Q., Hu, H., Lin, L. and Chen, Y.W. (2019) Deep Learning-Based Radiomics Models for Early Recurrence Prediction of Hepatocel-lular Carcinoma with Multi-Phase CT Images and Clinical Data. Annual International Conference of the IEEE Engineer-ing in Medicine and Biology Society, Berlin, 23-27 July 2019, 4881-4884.
https://doi.org/10.1109/EMBC.2019.8856356
[35] Zhou, Q., Zhou, Z., Chen, C., Fan, G., Chen, G., Heng, H., Ji, J. and Dai, Y. (2019) Grading of Hepatocellular Carcinoma Using 3D SE-DenseNet in Dynamic Enhanced MR Images. Computers in Biology and Medicine, 107, 47-57.
https://doi.org/10.1016/j.compbiomed.2019.01.026
[36] Yasaka, K., Akai, H., Abe, O. and Kiryu, S. (2018) Deep Learning with Convolutional Neural Network for Differentiation of Liver Masses at Dynamic Contrast-Enhanced CT: A Preliminary Study. Radiology, 286, 887-896.
https://doi.org/10.1148/radiol.2017170706
[37] Jiang, Y.Q., Cao, S.E., Cao, S., Chen, J.N., Wang, G.Y., Shi, W.Q., Deng, Y.N., et al. (2021) Preoperative Identification of Microvascular Invasion in Hepatocellular Carcinoma by XGBoost and Deep Learning. Journal of Cancer Research and Clinical Oncology, 147, 821-833.
https://doi.org/10.1007/s00432-020-03366-9
[38] Liu, S.C., Lai, J., Huang, J.Y., Cho, C.F., Lee, P.H., Lu, M.H., Yeh, C.C., Yu, J. and Lin, W.C. (2021) Predicting Microvascular Invasion in Hepatocellular Carcinoma: A Deep Learn-ing Model Validated across Hospitals. Cancer Imaging: The Official Publication of the International Cancer Imaging Society, 21, Article No. 56.
https://doi.org/10.1186/s40644-021-00425-3
[39] Vorontsov, E., Molchanov, P., Gazda, M., Beckham, C., et al. (2022) Towards Annotation-Efficient Segmentation via Image-to-Image Translation. Medical Image Analysis, 82, Article ID: 102624.
https://doi.org/10.1016/j.media.2022.102624
[40] Yang, Z. and Li, S. (2022) Dual-Path Network for Liver and Tumor Segmentation in CT Images Using Swin Transformer Encoding Approach. Current Medical Imaging.
[41] Pettit, R.W., Marlatt, B.B., Corr, S.J., Havelka, J. and Rana, A. (2022) nnU-Net Deep Learning Method for Segmenting Pa-renchyma and Determining Liver Volume from Computed Tomography Images. Annals of Surgery Open, 3, e155.
https://doi.org/10.1097/AS9.0000000000000155
[42] Guo, X., Schwartz, L.H. and Zhao, B. (2019) Automatic Liver Segmentation by Integrating Fully Convolutional Networks into Active Contour Models. Medical Physics, 46, 4455-4469.
https://doi.org/10.1002/mp.13735
[43] Wang, L., Wu, M., Li, R., Xu, X., Zhu, C. and Feng, X. (2022) MVI-Mind: A Novel Deep-Learning Strategy Using Computed Tomography (CT)-Based Radiomics for End-to-End High Efficiency Prediction of Microvascular Invasion in Hepatocellular Carcinoma. Cancers (Basel), 14, Article No. 2956.
https://doi.org/10.3390/cancers14122956