暂无相关信息!
[1]FERLAY J,COLOMBET M,SOERJOMATARAM I,et al.Cancer incidence and mortality patterns in Europe:estimates for 40 countries and 25 major cancers in 2018 [J].Eur J Cancer,2018,103:356-387.
[2]CAPITANIO U,BENSALAH K,BEX A,et al.Epidemiology of renal cell carcinoma [J].Eur Urol,2019,75(1):74-84.
[3]BUKAVINA L,BENSALAH K,BRAY F,et alEpidemiology of renal cell carcinoma:2022 update [J].Eur Urol,2022,82(5):529-542.
[4]SANCHEZ A,FELDMAN AS,HAKIMI AA.Currentmanagement of small renal masses,including patient selectionrenal tumor biopsy,active surveillance,and thermal ablation [J].J Clin Oncol,2018,36(36):3591-3600.
[5]ZHU YH,WANG X,ZHANG J,et al.Low enhancementon multiphase contrast-enhanced CT images:an independentpredictor of the presence of high tumor grade of clear cel renal cell carcinoma [J].AjR Am J Roentgenol,2014,203(3):W295-W300.
[6] VARGAS HA,DELANEY HG,DELAPPE EM,et al.Multiphasic contrast-enhanced MRI:single-slice versus volumetric quantification of tumor enhancement for the assessment of renal clear-cell carcinoma fuhrman grade [J].J Magn Reson Imaging,2013,37(5):1160-1167.
[7]KIM C,CHOI HJ,CHO KS.Diagnostic value of multidetector computed tomography for renal sinus fat invasion in renal cell carcinoma patients [J].Eur J Radiol, 2014,83(6):914-918.
[8]COTTA BH,MEAGHER MF,BRADSHAW A,et al.Percutaneous renal mass biopsy:historical perspective,current status,and future considerations [J].Expert REVAnticancer Ther,2019,19(4):301-308.
[9]GILLIES RJ,KINAHAN PE,HRICAK H.Radiomics:images are more than pictures [J].They Are Data.Radiology,2016,278(2):563-577.
[10]LAMBIN P,RIOS-VELAZQUEZ E,LEIJENAAR R,et al.Radiomics:extracting more information from medicalimages using advanced feature analysis [J].Eur J Cancer,2012,48(4):441-446.
[11]YIP SS,AERTS HJ.Applications and limitations ofradiomics [J].Phys Med Biol,2016,61(13):R150-R166.
[12]MCBEE MP,AWAN OA,COLUCCI AT,et al.Deeplearning in radiology [J].Acad Radiol,2018,25(11):1472-1480.
[13]HE KM,ZHANG XY,REN SQ,et al.Deep residuallearning for image recognition [A].Proceedings of the2016 IEEE Conference on Computer Vision and PatternRecognition(CVPR)[C].Las Vegas,NV,USA:IEEE,2016:770-778
[14]HELLER N,ISENSEE F,MAIER-HEIN KH,et al.The state of the art in kidney and kidney tumor segmentationin contrast-enhanced CT imaging:results of the KiTS19challenge [J].Med Image Anal,2021,67:101821.
[15]余苔痕,基于CT的深度学习预测肾透明细胞癌病理分级及分子分型的研究[D].广州:南方医科大学,2022.
[16]RAO PK,CHATTERJEE S,JANARDHAN M,et al.Optimizing inference distribution for efficient kidney tumorsegmentation using a UNet-PWP deep-learning model withXAI on CT scan images [J].Diagnostics (Basel),2023,13(20):3244.
[17]KLONTZAS ME,KALARAKIS G,KOLTSAKIS E,et al.Convolutional neural networks for the differentiation betweenbenign and malignant renal tumors with a multicenter internationalcomputed tomography dataset [J].Insights Imaging,2024,15(1):26.
[18]NIKPANAH M,XU Z,JIN D,et al.A deep-learningbased artificial intelligence(AI)approach for differentiationof clear cell renal cellcarcinoma from oncocytoma on multi-phasic MRI [J].Clin Imaging,2021,77:291-298.
[19]PEDERSEN M,ANDERSEN MB,CHRISTIANSEN H,et al.Classification of renal tumour using convolutional neural networks to detect oncocytoma [J].Eur J Radiol,
2020,133:109343
[20]XI IL,ZHAO Y,WANG R,et al.Deep learning todistinguish benign from malignant renal lesions based onroutine MR imaging [J].Clin Cancer Res,2020,26(8):1944-1952.
[21]BAGHDADI A,ALDHAAM NA,ELSAYED AS,et alAutomated differentiation of benign renal oncocytoma andchromophobe renal cell carcinoma on computed tomographyusing deep learning [J].BJU Int,2020,125(4):553-560.
[22]ZUO T,ZHENG Y,HE L,et al.Automated classificationof papillary renal cell carcinoma and chromophobe renal cellcarcinoma based on a small computed tomography imagingdataset using deep learning [J].Front Oncol,2021,11:746750.
[23]HAN S,HWANG SI,LEE HJ.The classification ofrenal cancer in 3-phase CT images using a deep learningmethod [J].J Digit Imaging,2019,32(4):638-643.
[24]CHEN S,SONG D,CHEN L,et al.Artificialintelligence-based non-invasive tumor segmentation,gradestratification and prognosis prediction for clear-cell renal-cell carcinoma [J].Precis Clin Med,2023,6(3):pbad019.
[25]YANG M,HE X,XU L,et al.CT-based transformermodel for non-invasively predicting the fuhrman nucleargrade of clear cell renal cell carcinoma [J].Front Oncol 2022,12:961779.
[26] XU L,YANG C,ZHANG F,et al.Deep learning usingCT images to grade clear cell renal cell carcinoma:developmentand validation of a prediction model [J].Cancers(Basel),2022,14(11):2574.
[27]ZHAO Y,CHANG M,WANG R,et al.Deep learning based on MRI for differentiation of low-and high-grade inlow-stage renal cell carcinoma [J].J Magn Reson Imaging.2020,52(5):1542-1549.
[28]FICARRA V,GALFANO A,MANCINI M,et al.TNMstaging system for renal-cell carcinoma:current status andfuture perspectives [J].Lancet Oncol,2007,8(6):554-558
[29]DEMIRJIAN NL,VARGHESE BA,CEN SY,et al.CT-based radiomics stratification of tumor grade and TNMstage of clear cell renal cell carcinoma [J].Eur Radiol,2022,32(4):2552-2563.
[30]YANG L,GAO L,AREFAN D,et al.A CT-basedradiomics model for predicting renal capsule invasion in renalcell carcinoma [J].BMC Med Imaging,2022,22(1):15.
[31]NIE P,YANG G,WANG Y,et al.A CT-based deeplearning radiomics nomogram outperforms the existing prognosticmodels for outcome prediction in clear cell renal cellcarcinoma:a multicenter study [J].Eur Radiol,2023,33(12):8858-8868
[32]WANG S,ZHU C,JIN Y,et al.A multi-model basedon radiogenomics and deep learning techniques associatedwith histological grade and survival in clear cell renal cellcarcinoma [J].Insights Imaging,2023,14(1):207.
[33]KHENE ZE,MATHIEU R,PEYRONNET B,et al.Radiomics can predict tumour response in patients treatedwith nivolumab for a metastatic renal cell carcinoma:anartificial intelligence concept [J].World J Urol,2021,39(9): 3707-3709.
[34]HAIDER MA,VOSOUGH A,KHALVATI F,et al.CT texture analysis:a potential tool for prediction of survival in patients with metastatic clear cell carcinoma treated with sunitinib [J].Cancer Imaging,2017,17(1):4.
吕定阳,双卫兵.影像组学与深度学习在肾癌精准诊疗中的研究进展[J].泌尿外科杂志(电子版),2024,16(02):50-54.DOI:10.20020/j.CNKI.1674-7410.2024.02.09.
暂无相关信息!
肾癌是泌尿系统中常见恶性肿瘤之一,占所有癌症的3%[1-2]。随着影像技术的进步和社会生活水平的提高,肾癌的诊断率正逐年增长。2020年,全球约有431288例肾癌新发病例[3]。影像学检查在肾癌的诊断、疗效评价及预后预测等方面均发挥重要作用。传统电子计算机断层扫描(computed tomography,CT)、磁共振成像(magnetic resonance imaging, MRI)等检查可显示肿瘤的形态学特征,如部位、大小、密度或信号特点及强化程度等,但诊断灵敏度和特异度有限,难以将嗜酸细胞瘤、乏脂性肾错构瘤与肾癌相鉴别[4]。病理分级和TNM分期是肾癌预后的重要预测因素,不同影像学研究结果间存在争议[5-6]。此外,CT和MRI对肾癌局部浸润(pT3期)的鉴别能力有限[7]。肾癌诊断的金标准为穿刺活检和肿瘤切除术后的组织病理学分析,为患者预后提供重要信息。组织病理标本的获取为侵入性方法,且由于肿瘤的异质性,活检不能表征肿瘤的整体情况,有低估肿瘤病理分级的风险[8]。随着人工智能的发展,深度学习与影像组学已经广泛应用于肿瘤的预测评估。影像组学能够考虑整个肿瘤区域,从而更好地表征肿瘤异质性[9]。通过挖掘CT和MRI等影像图像的高通量信息,利用深度学习构建准确、可重复性的预测模型,可实现术前对肾癌患者诊断、组织学分型、病理分级、疗效的评估及生存预测,有助于制定个体化诊疗方案。本文将从影像组学与深度学习及其在肾癌中的应用进行综述,以期为肾癌的临床诊疗策略提供依据。
1 影像组学
2 深度学习
3 深度学习与影像组学在肾癌诊疗中的应用
3.1 肾脏肿瘤的自动分割
3.2 肾癌与肾脏良性肿瘤鉴别
3.3 肾癌组织学亚型鉴别
3.4 预测肾癌病理分级
3.5 肾癌预后预测
3.6 疗效评估
4 总结与展望
影像组学和深度学习已逐渐应用于肾脏肿瘤的自动分割、肾癌与肾脏良性肿瘤鉴别、肾癌组织学亚型鉴别、肾癌病理分级和肾癌预后预测等方面。但其尚处于临床前研究阶段,至临床应用还需得到充分的验证,如大样本的前瞻性研究或独立的外部验证数据集。影像组学和深度学习模型的价值,在于其表现能否超越传统影像标志物和现有预测模型,或能否提升放射科医师的诊断准确性,只有满足这些,才能奠定临床转化的基础条件。未来还需通过大规模、多中心研究来验证影像组学和深度学习在肾癌诊疗中的应用价值。随着数据不断积累及技术持续进步,多模态、多组学的研究将成为重点,旨在解决临床核心问题,满足精准医学的需求。
暂无相关信息!
暂无相关信息!
作者相关文章