CLC number:
On-line Access: 2024-08-27
Received: 2023-10-17
Revision Accepted: 2024-05-08
Crosschecked: 2024-01-02
Cited: 0
Clicked: 1015
Zhikun LIU, Yichao WU, Abid Ali KHAN, Lun LU, Jianguo WANG, Jun CHEN, Ningyang JIA, Shusen ZHENG, Xiao XU. Deep learning-based radiomics allows for a more accurate assessment of sarcopenia as a prognostic factor in hepatocellular carcinoma[J]. Journal of Zhejiang University Science B,in press.Frontiers of Information Technology & Electronic Engineering,in press.https://doi.org/10.1631/jzus.B2300363 @article{title="Deep learning-based radiomics allows for a more accurate assessment of sarcopenia as a prognostic factor in hepatocellular carcinoma", %0 Journal Article TY - JOUR
基于影像组学的深度学习评估肌少症对肝癌切除和移植患者预后的影响1浙江省肿瘤融合与智能医学实验室,浙江大学医学院附属杭州市第一人民医院肝胆胰外科,中国杭州市,310006 2浙江大学医学院,中国杭州市,310058 3海军医科大学东方肝胆外科医院放射科,中国上海市,200438 4浙江大学医学院附属第一附属医院肝胆胰外科,中国杭州市,310003 5国家卫健委多器官联合移植重点实验室,中国杭州市,310003 6树兰(杭州)医院肝胆胰外科,中国杭州市,310022 摘要:肌少症是指因持续骨骼肌含量流失、强度和功能下降引起的综合征,且与包括肝细胞癌(HCC)在内的肿瘤患者预后密切相关。目前该病的检测手段局限且无统一标准。本文旨在利用基于影像组学的深度学习(DL)技术评估肌少症,用于肝癌患者行肝脏部分切除术或肝移植术的预后预测。本研究回顾性纳入浙大一院肝癌手术切除492例(训练集+内部验证集)与肝癌肝移植173例患者(外部LT验证集),东方肝胆医院肝癌切除患者161例(外部验证集),并收集患者术前一个月内的腹部计算机断层扫描(CT)平扫期影像与临床资料;单中心肝切除术组入组患者按7:3随机分为训练集和内部验证集(训练集345例,验证集147例),肝移植组及第二中心肝癌切除组作为外部验证集,经训练集建立预测模型,并利用内部和外部验证集验证预测模型的预测性能;对训练集患者CT图像中第3腰椎骨(L3)层面的骨骼肌(SM)及腰大肌(PM)轮廓进行人工勾画;抽提SM与PM影像组学特征,随后利用自编码器(AutoEncoder)压缩特征,TFDeepSurv生存分析网络构建DL预后预测模型,预测HCC术后无瘤生存率(RFS)与总体生存时间(OS);最后计算时间依赖性受试者工作特征曲线(ROC)的曲线下面积(AUC)和一致性指数(C-index),采用应用净重新分类改善指数(NRI)和临床决策曲线(DCA)评价模型预测性能。最终从勾画的CT图像L3层面的SM及PM中采集相应肌肉中1343个影像组学特征。经AutoEncoder将此高阶影像组学特征降维至100个特征。运用TFDeepSurv生存分析网络完成DL预测模型的构建,将HCC患者根据预后的差异分为高危组和低危组,高危组HCC患者行肝部分切除手术后预后显著低于低危组患者。此外,通过Kaplan-Meier生存曲线分析等方法证实DL模型在内部及外部验证集、外部LT验证集中均可对肝癌患者术后的预后进行准确预测,一致性指数分别达0.775和0.613。NRI和DCA同样显示DL模型具有较高的预测性能。本研究创新性地提出了基于影像组学的DL技术构建的预后预测模型;该模型可在术前对肝癌手术切除和肝移植术后的生存风险进行个体化预测,从而实现对肝癌患者OS的早期预判,有助于制定合理的临床决策和指导临床实践。 关键词组: Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article
Reference[1]BiWL, HosnyA, SchabathMB, et al., 2019. Artificial intelligence in cancer imaging: clinical challenges and applications. CA Cancer J Clin, 69(2):127-157. [2]CareyEJ, LaiJC, WangCW, et al., 2017. A multicenter study to define sarcopenia in patients with end-stage liver disease. Liver Transpl, 23(5):625-633. [3]CareyEJ, LaiJC, SonnendayC, et al., 2019. A North American expert opinion statement on sarcopenia in liver transplantation. Hepatology, 70(5):1816-1829. [4]ChenXD, ChenWJ, HuangZX, et al., 2022. Establish a new diagnosis of sarcopenia based on extracted radiomic features to predict prognosis of patients with gastric cancer. Front Nutr, 9:850929. [5]Cruz-JentoftAJ, SayerAA, 2019. Sarcopenia. Lancet, 393(10191):2636-2646. [6]Cruz-JentoftAJ, BahatG, BauerJ, et al., 2019. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing, 48(4):601. [7]EsserH, ReschT, PammingerM, et al., 2019. Preoperative assessment of muscle mass using computerized tomography scans to predict outcomes following orthotopic liver transplantation. Transplantation, 103(12):2506-2514. [8]FornerA, ReigM, BruixJ, 2018. Hepatocellular carcinoma. Lancet, 391(10127):1301-1314. [9]FujiwaraN, NakagawaH, KudoY, et al., 2015. Sarcopenia, intramuscular fat deposition, and visceral adiposity independently predict the outcomes of hepatocellular carcinoma. J Hepatol, 63(1):131-140. [10]GolseN, BucurPO, CiacioO, et al., 2017. A new definition of sarcopenia in patients with cirrhosis undergoing liver transplantation. Liver Transpl, 23(2):143-154. [11]HamaguchiY, KaidoT, OkumuraS, et al., 2016. Proposal for new diagnostic criteria for low skeletal muscle mass based on computed tomography imaging in Asian adults. Nutrition, 32(11-12):1200-1205. [12]HeZQ, SheXM, LiuZY, et al., 2023. Advances in post-operative prognostic models for hepatocellular carcinoma. J Zhejiang Univ-Sci B (Biomed & Biotechnol), 24(3):191-206. [13]JinJY, YaoZ, ZhangT, et al., 2021. Deep learning radiomics model accurately predicts hepatocellular carcinoma occurrence in chronic hepatitis b patients: a five-year follow-up. Am J Cancer Res, 11(2):576-589. [14]KimYJ, 2021. Machine learning models for sarcopenia identification based on radiomic features of muscles in computed tomography. Int J Environ Res Public Health, 18(16):8710. [15]van VugtJLA, LevolgerS, de BruinRWF, et al., 2016. Systematic review and meta-analysis of the impact of computed tomography-assessed skeletal muscle mass on outcome in patients awaiting or undergoing liver transplantation. Am J Transpl, 16(8):2277-2292. [16]VoronT, TselikasL, PietraszD, et al., 2015. Sarcopenia impacts on short- and long-term results of hepatectomy for hepatocellular carcinoma. Ann Surg, 261(6):1173-1183. [17]YooT, LoWD, EvansDC, 2017. Computed tomography measured psoas density predicts outcomes in trauma. Surgery, 162(2):377-384. Journal of Zhejiang University-SCIENCE, 38 Zheda Road, Hangzhou
310027, China
Tel: +86-571-87952783; E-mail: cjzhang@zju.edu.cn Copyright © 2000 - 2024 Journal of Zhejiang University-SCIENCE |
Open peer comments: Debate/Discuss/Question/Opinion
<1>