摘 要：目前关于非小细胞肺癌（NSCLC）患者生存分析的研究已经有很多，但是大多数都是在医生勾画出肿瘤的基础上进行影像组学特征提取，其次结合临床以及治疗前患者的肿瘤 PET/CT 图像特征进行生存分析的研究。在无医生勾画肿瘤的基础上，采用深度学习的方法，基于患者治疗前后 FDG-PET 是否可以对局部晚期 NSCLC 患者进行生存分析。在采用治疗前和治疗后 FDG-PET 时，基于 3D 卷积神经网络（3D CNN）的深度生存模型的一致性指数（C-index）为 0.67。研究表明，同时使用治疗前后 PDG-PET 进行阅片可以预测出患者的风险概率。
关键词：非小细胞肺癌；治疗前后 PDG-PET；3D 卷积神经网络；生存分析
DOI:10.19850/j.cnki.2096-4706.2023.04.028
中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2023）04-0109-05
Survival Prediction of Patients with Locally Advanced NSCLC Based on 3D CNN
MA Ya'nan, SONG Yue, HAO Tianyu
(Lanzhou University of Finance and Economics, Lanzhou 730020, China)
Abstract: There have been many studies on the survival analysis of patients with non-small cell lung cancer (NSCLC). However, most of the studies are based on the extraction of tumor radiomics features based on the tumour label outlined by the physician, followed by a combination of clinical and pre-treatment PET/CT image features of the patient for survival analysis. Survival analysis of patients with locally advanced NSCLC based on whether pre-treatment and post-treatment FDG-PET can be performed by using a deep learning approach without the basis of tumors label of the physician. The consistency index (C-index) of the deep survival model based on 3D CNN is 0.67 when using pre-treatment and post-treatment FDG-PET. The study shows that simultaneous reading with pre-treatment and post-treatment PDG-PET can predict the risk probability of patient.
Keywords: non-small cell lung cancer; pre-treatment and post-treatment PDG-PET; 3D CNN; survival analysis
参考文献：
[1] SIEGEL R L，MILLER K D，FUCHS H E，et al. Cancer Statistics [J].CA：A Cancer Journal for Clinicians.2022，72（1）：7–33.
[2] 中华医学会肿瘤学分会，中华医学会杂志社 . 中华医学会肿瘤学分会肺癌临床诊疗指南：2021 版 [J]. 中华肿瘤杂志，2021，43（6）：591-621.
[3] SHIARLI A-M，PATEL P，MCDONALD F. Management of Locally Advanced Non-Small Cell Lung Cancer [M]//JANES S M. Encyclopedia of Respiratory Medicine：Second Edition.Pittsburgh： Academic Press，2022：810-824.
[4] TRUONG M T，VISWANATHAN C，ERASMUS J J. Positron Emission Tomography/Computed Tomography in Lung Cancer Staging，Prognosis，and Assessment of Therapeutic Response [J]. Journal of Thoracic Imaging，2011，26（2）：132-46.
[5] COX D R. Regression Models and Life-Tables [M]//KOTZ S， JOHNSON N L. Breakthroughs in Statistics Volume II Methodology and Diitribution.Berlin：Springer，1992：527-541.
[6] HINTON G. Deep Learning—a Technology with the Potential to Transform Health Care [J].JAMA，2018，320（11）：1101-1102.
[7] LECUN Y，BENGIO Y，HINTON G. Deep learning [J]. Nature，2015，521（28）：436-444.
[8] KATZMAN J L，SHAHAM U，CLONINGER A，et al. Deepsurv：Personalized Treatment Recommender System Using a Cox Proportional Hazards Deep Neural Network [J/OL].arXiv:1606.00931 [stat.ML].[2022-08-03].https://arxiv.org/abs/1606.00931.
[9] VIZCARRA J，PLACE R，TONG L，et al. Fusion in Breast Cancer Histology Classification [C]//the 10th ACM International Conference on Bioinformatics，Computational Biology and Health Informatics.Niagara Falls：ACM，2019：485-493.
[10] NISHIKAWA R M，GIGER M L，DOI K，et al. Effect of Case Selection on the Performance of Computer-Aided Detection Schemes [J].Medical physics，1994，21（2）：265-269.
[11] RODRIGUEZ-RUIZ A，TEUWEN J，CHUNG K，et al. Pectoral Muscle Segmentation in Breast Tomosynthesis with Deep Learning [C]//Medical Imaging 2018：Computer-Aided Diagnosis. Houston：SPIE，2018：10575.
[12] WOO S，PARK J，LEE J-Y，et al. CBAM：Convolutional Block Attention Module [J/OL].arXiv:1807.06521 [cs.CV].[2022-08-06].https://arxiv.org/abs/1807.06521v2.
[13] YAO J W，ZHU X L，JONNAGADDALA J，et al. Whole Slide Images Based Cancer Survival Prediction Using Attention Guided Deep Multiple Instance Learning Networks [J/OL].arXiv:2009.11169[eess.IV].[2022-08-09].https://arxiv.org/abs/2009.11169.
作者简介：马亚楠（1996—），女，汉族，河南郑州人，硕士在读，研究方向：信息管理与信息系统；宋玥（1997—），女，汉族，山西吕梁人，硕士在读，研究方向：信息管理与信息系统；郝天宇（2001—），男，汉族，湖北荆门人，本科在读，研究方向：信息管理与信息系统。