
浏览全部资源
扫码关注微信
1.陆军特色医学中心(第三军医大学大坪医院)肿瘤科,重庆
2.陆军军医大学(第三军医大学)生物医学工程与影像医学系数字医学教研室,重庆
3.渝粤病理科学研究中心,重庆
4.金凤实验室,重庆
李梦侠,E-mail: limengxia@tmmu.edu.cn
收稿日期:2024-10-29,
修回日期:2025-03-19,
纸质出版日期:2025-07-30
移动端阅览
张地,吴毅,徐瑜,等.基于机器学习的多模态数据建立Ⅰ期非小细胞肺癌患者术后复发预测模型[J].陆军军医大学学报,2025,47(14):1602-1611.DOI:10.16016/j.2097-0927.202410117.
ZHANG D,WU Y,XU Y,et al.Development of a postoperative recurrence prediction model for stage Ⅰ non-small cell lung cancer patients using multimodal data based on machine learning[J].J Army Med Univ,2025,47(14):1602-1611.DOI:10.16016/j.2097-0927.202410117.
张地,吴毅,徐瑜,等.基于机器学习的多模态数据建立Ⅰ期非小细胞肺癌患者术后复发预测模型[J].陆军军医大学学报,2025,47(14):1602-1611.DOI:10.16016/j.2097-0927.202410117. DOI:
ZHANG D,WU Y,XU Y,et al.Development of a postoperative recurrence prediction model for stage Ⅰ non-small cell lung cancer patients using multimodal data based on machine learning[J].J Army Med Univ,2025,47(14):1602-1611.DOI:10.16016/j.2097-0927.202410117. DOI:
目的
2
基于术前胸部CT影像组学特征联合临床信息构建机器学习模型,用于预测手术切除的Ⅰ期非小细胞肺癌患者术后5年复发风险。
方法
2
回顾性纳入2014年1月至2019年12月陆军特色医学中心经手术病理确诊的Ⅰ期非小细胞肺癌患者217例(入组标准筛选自778例初筛病例),其中术后5年内复发53例,未复发164例。将病例按8∶2比例随机分为训练集(173例)和验证集(44例)。以肺癌原发灶主体作为感兴趣区域(ROI)提取影像组学特征,建立影像模型;同时基于患者人口学资料及术前检验等指标建立临床模型。进一步融合两类特征构建融合模型,并通过模型性能比较筛选最优预测模型。
结果
2
本研究通过平扫CT影像特征筛选,最终纳入7个影像组学特征用于影像模型构建。在6种机器学习模型中,自适应增强算法(adaptive boosting, Adaboost)的训练集曲线下面积(AUC)为0.866(95%
CI
0.808~0.923),准确度和特异度分别为0.832、0.884;验证集AUC为0.806(95%
CI
0.630~0.983),准确度和特异度分别为0.795、0.971,综合比较预测性能最好。基于单因素及多因素Logistic回归分析筛选了4个临床特征参与构建临床模型,在训练集中AUC为0.874(95%
CI
0.821~0.928),准确度和特异度分别为0.827、0.891;验证集中AUC为0.813(95%
CI
0.677~0.948),准确度和特异度分别为0.636、0.600。进一步通过特征级融合策略整合影像组学与临床特征后,融合模型在训练集AUC达到0.953(95%
CI
0.924~0.983
),准确度和特异度分别为0.884、0.860;在验证集AUC达到0.852(95%
CI
0.729~0.976),准确度和特异度分别为0.682、0.629,预测效能得到进一步提升。
结论
2
融合术前CT影像组学特征与临床风险因素构建的联合模型,可为评估Ⅰ期非小细胞肺癌患者术后5年内的复发风险提供依据。
Objective To develop a machine learning model integrating preoperative chest CT radiomic features with clinical data for predicting 5-year postoperative recurrence risk in stage Ⅰ non-small cell lung cancer (NSCLC) patients undergoing surgical resection. Methods A total of 217 patients with pathologically confirmed stage Ⅰ NSCLC (selected from 778 initially screened cases based on our inclusion and exclusion criteria) treated in Army Medical Center of PLA between January 2014 and December 2019 were retrospectively enrolled
including 53 recurrence cases and 164 non-recurrence cases within 5-year follow-up. They were randomly divided into a training set (
n
=173) and a validation set (
n
=44) in a ratio of 8:2. Radiomic models were established based on extracted features from tumor-dominant regions of interest (ROI) on CT images
while clinical models were developed using demographic characteristics and preoperative laboratory examinations. A combined model was further constructed by integrating both feature sets
and model performance was compared to identify the optimal predictive model.Results This study screened the features from non-contrast CT images and ultimately selected 7 radiomic features for constructing radiomic model. Among 6 machine learning algorithms
the adaptive boosting (Adaboost) model demonstrated the best overall predictive performance
with an area under the curve (AUC) of 0.866 (95%
CI
: 0.808~0.923; accuracy: 0.832
specificity: 0.884) in the training set and of 0.806 (95%
CI
: 0.630~0.983; accuracy: 0.795
specificity: 0.971) in the validation set. Univariate and multivariate logistic regression analyses identified 4 clinical features for clinical m
odel construction. The clinical model achieved an AUC value of 0.874 (95%
CI
: 0.821~0.928; accuracy: 0.827
specificity: 0.891) in the training set and 0.813 (95%
CI
: 0.677~0.948; accuracy: 0.636
specificity: 0.600) in the validation set. By integrating the 7 radiomic features and 4 clinical features using a feature-level fusion strategy
the combined model exhibited further improved predictive performance
with an AUC value of 0.953 (95%
CI
: 0.924~0.983; accuracy: 0.884
specificity: 0.860) and 0.852 (95%
CI
: 0.729~0.976; accuracy: 0.682
specificity: 0.629)
respectively in the training set and the validation set. Conclusion The combined model integrating preoperative CT radiomic features with clinical risk factors may provide an evidence-based framework for evaluating 5-year postoperative recurrence risk in stage Ⅰ NSCLC patients.
SUNG H , FERLAY J , SIEGEL R L , et al . Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries [J]. CA Cancer J Clin , 2021 , 71 ( 3 ): 209 - 249 . DOI: 10.3322/caac.21660 http://dx.doi.org/10.3322/caac.21660 .
中华人民共和国国家卫生健康委员会 . 原发性肺癌诊疗指南(2022年版) [J]. 中国合理用药探索 , 2022 , 19 ( 9 ): 1 - 28 . DOI: 10.3969/j.issn.2096-3327.2022.09.002 http://dx.doi.org/10.3969/j.issn.2096-3327.2022.09.002 .
National Health Commission of the People’s Republic of China . Guidelines for the diagnosis and treatment of primary lung cancer (2022 edition) [J]. Chin J Ration Drug Use , 2022 , 19 ( 9 ): 1 - 28 . DOI: 10.3969/j.issn.2096-3327.2022.09.002 http://dx.doi.org/10.3969/j.issn.2096-3327.2022.09.002 .
SINGAREDDY A , FLANAGAN M E , SAMSON P P , et al . Trends in stage I lung cancer [J]. Clin Lung Cancer , 2023 , 24 ( 2 ): 114 - 119 . DOI: 10.1016/j.cllc.2022.11.005 http://dx.doi.org/10.1016/j.cllc.2022.11.005 .
RUAN Z G , ZHUO X , XU C Y . Diagnosis, treatment, and prognosis of stage IB non-small cell lung cancer with visceral pleural invasion [J]. Front Oncol , 2024 , 13 : 1310471 . DOI: 10.3389/fonc.2023.1310471 http://dx.doi.org/10.3389/fonc.2023.1310471 .
SIEGEL R L , MILLER K D , WAGLE N S , et al . Cancer statistics, 2023 [J]. CA Cancer J Clini , 2023 , 73 ( 1 ): 17 - 48 . DOI: 10.3322/caac.21763 http://dx.doi.org/10.3322/caac.21763 .
AERTS H J W L . The potential of radiomic-based phenotyping in precision medicine: a review [J]. JAMA Oncol , 2016 , 2 ( 12 ): 1636 - 1642 . DOI: 10.1001/jamaoncol.2016.2631 http://dx.doi.org/10.1001/jamaoncol.2016.2631 .
WU G Y , WOODRUFF H C , SHEN J , et al . Diagnosis of invasive lung adenocarcinoma based on chest CT radiomic features of part-solid pulmonary nodules: a multicenter study [J]. Radiology , 2020 , 297 ( 2 ): 451 - 458 . DOI: 10.1148/radiol.2020192431 http://dx.doi.org/10.1148/radiol.2020192431 .
赵媛 , 王洪峰 , 赵林 , 等 . 能谱CT影像组学非小细胞肺癌淋巴结转移的预测模型构建 [J]. 中国临床医学影像杂志 , 2024 , 35 ( 9 ): 628 - 632 . DOI: 10.12117/jccmi.2024.09.005 http://dx.doi.org/10.12117/jccmi.2024.09.005 .
ZHAO Y , WANG H F , ZHAO L , et al . Construction of a predictive model for lymph node metastasis in non-small cell lung cancer based on spectral CT radiomics [J]. J China Clin Med Imag , 2024 , 35 ( 9 ): 628 - 632 . DOI: 10.12117/jccmi.2024.09.005 http://dx.doi.org/10.12117/jccmi.2024.09.005 .
SAAD M B , SHOWKATIAN E , AL-TASHI Q , et al . I-SABR-SELECT: a radiomics-based model for personalized immunotherapy for early-stage non-small cell lung cancer [J]. Int J Radiat Oncol , 2024 , 120 ( 2 ): S42 - S43 . DOI: 10.1016/j.ijrobp.2024.07.064 http://dx.doi.org/10.1016/j.ijrobp.2024.07.064 .
AMINI M , NAZARI M , SHIRI I , et al . Multi-level multi-modality (PET and CT) fusion radiomics: prognostic modeling for non-small cell lung carcinoma [J]. Phys Med Biol , 2021 , 66 ( 20 ): 205017 . DOI: 10.1088/1361-6560/ac287d http://dx.doi.org/10.1088/1361-6560/ac287d .
王述平 . 基于MRI影像组学的肝癌病型及分化等级预测研究 [D]. 保定 : 河北大学 , 2023 . DOI: 10.27103/d.cnki.ghebu.2023.000672 http://dx.doi.org/10.27103/d.cnki.ghebu.2023.000672 .
WANG S P . Prediction of liver cancer type anddifferentiation grade based on MRI radiomics [D]. Baoding : Hebei University , 2023 . DOI: 10.27103/d.cnki.ghebu.2023.000672 http://dx.doi.org/10.27103/d.cnki.ghebu.2023.000672 .
李旭 . CT影像组学在食管癌淋巴结转移及浸润深度预测中的研究 [D]. 济南 : 山东大学 , 2023 .
LI X . A computed tomography-based clinical-radiomics model for prediction of lymph node metastasis and Infiltration depth in esophageal carcinoma [D]. Jinan : Shandong University , 2023 .
陆亮 , 徐圆 . 影像组学联合CT特征对周围型肺腺癌及鳞癌的鉴别价值研究 [J]. 医疗卫生装备 , 2021 , 42 ( 10 ): 48 - 52, 63 . DOI: 10.19745/j.1003-8868.2021211 http://dx.doi.org/10.19745/j.1003-8868.2021211 .
LU L , XU Y . Diagnostic value of radiomics combined with CT features for differentiating peripheral lung adenocarcinoma from squamous cell carcinoma [J]. Chin Med Equip J , 2021 , 42 ( 10 ): 48 - 52, 63 . DOI: 10.19745/j.1003-8868.2021211 http://dx.doi.org/10.19745/j.1003-8868.2021211 .
师庆科 , 李楠 , 叶枫 . 机器学习在临床研究中的应用进展 [J]. 中国数字医学 , 2025 , 20 ( 3 ): 1 - 10 . DOI: 10.3969/j.issn.1673-7571.2025.03.001 http://dx.doi.org/10.3969/j.issn.1673-7571.2025.03.001 .
SHI Q K , LI N , YE F . Advancements in the application of machine learning for clinical research [J]. China Digit Med , 2025 , 20 ( 3 ): 1 - 10 . DOI: 10.3969/j.issn.1673-7571.2025.03.001 http://dx.doi.org/10.3969/j.issn.1673-7571.2025.03.001 .
SONG F , SONG X , FENG Y D , et al . Radiomics feature analysis and model research for predicting histopathological subtypes of non-small cell lung cancer on CT images: a multi-dataset study [J]. Med Phys , 2023 , 50 ( 7 ): 4351 - 4365 . DOI: 10.1002/mp.16233 http://dx.doi.org/10.1002/mp.16233 .
刘婧娟 , 刘炜 , 薛华丹 , 等 . 基于增强CT的影像组学定量特征模型在预测胃癌分期中的价值 [J]. 放射学实践 , 2022 , 37 ( 9 ): 1061 - 1067 . DOI: 10.13609/j.cnki.1000-0313.2022.09.002 http://dx.doi.org/10.13609/j.cnki.1000-0313.2022.09.002 .
LIU J J , LIU W , XUE H D , et al . The value of contrast enhanced CT-based radiomics analysis in predicting gastric cancer stage [J]. Radiol Pract , 2022 , 37 ( 9 ): 1061 - 1067 . DOI: 10.13609/j.cnki.1000-0313.2022.09.002 http://dx.doi.org/10.13609/j.cnki.1000-0313.2022.09.002 .
黄蓉 , 卢晓旭 , 孙学明 , 等 . 18 F-FDG PET/CT联合临床建立多模态影像组学模型预测局部晚期头颈部鳞癌对于新辅助免疫联合化学治疗的反应 [J]. 中国医学影像技术 , 2024 , 40 ( 10 ): 1493 - 1498 . DOI: 10.13929/j.issn.1003-3289.2024.10.008 http://dx.doi.org/10.13929/j.issn.1003-3289.2024.10.008 .
HUANG R , LU X X , SUN X M , et al . Multimodal models established combined 18 F-FDG PET/CT radiomics with clinical data for evaluating response of locally advanced head and neck squamous cell carcinoma to neoadjuvant immuno-chemotherapy [J]. Chin J Med Imag Technol , 2024 , 40 ( 10 ): 1493 - 1498 . DOI: 10.13929/j.issn.1003-3289.2024.10.008 http://dx.doi.org/10.13929/j.issn.1003-3289.2024.10.008 .
贺安晶 . 基于CT影像组学预测晚期肺腺癌靶向治疗短期疗效 [D]. 南昌 : 南昌大学 , 2024 . DOI: 10.27232/d.cnki.gnchu.2024.003659 http://dx.doi.org/10.27232/d.cnki.gnchu.2024.003659 .
HE A J . Prediction of short-term efficacy of targeted therapy for advanced lung adenocarcinoma based on CT radiomics [D]. Nanchang : Nanchang University , 2024 . DOI: 10.27232/d.cnki.gnchu.2024.003659 http://dx.doi.org/10.27232/d.cnki.gnchu.2024.003659 .
HARRIS H . The human alkaline phosphatases: what we know and what we don’t know [J]. Clin Chim Acta , 1990 , 186 ( 2 ): 133 - 150 . DOI: 10.1016/0009-8981(90)90031-m http://dx.doi.org/10.1016/0009-8981(90)90031-m .
ZHANG L , GONG Z X . Clinical characteristics and prognostic factors in bone metastases from lung cancer [J]. Med Sci Monit , 2017 , 23 : 4087 - 4094 . DOI: 10.12659/msm.902971 http://dx.doi.org/10.12659/msm.902971 .
CATASSI A , SERVENT D , PALEARI L , et al . Multiple roles of nicotine on cell proliferation and inhibition of apoptosis: implications on lung carcinogenesis [J]. Mutat Res , 2008 , 659 ( 3 ): 221 - 231 . DOI: 10.1016/j.mrrev.2008.04.002 http://dx.doi.org/10.1016/j.mrrev.2008.04.002 .
张凯 , 李树建 , 陈小乐 . 血浆D-二聚体水平在非小细胞肺癌患者预后评估中的价值 [J]. 医学综述 , 2022 , 28 ( 5 ): 1032 - 1035 . DOI: 10.3969/j.issn.1006-2084.2022.05.035 http://dx.doi.org/10.3969/j.issn.1006-2084.2022.05.035 .
ZHANG K , LI S J , CHEN X L . Value of plasma D-dimer level in prognostic assessment of patients with non-small cell lung cancer [J]. Med Recapitul , 2022 , 28 ( 5 ): 1032 - 1035 . DOI: 10.3969/j.issn.1006-2084.2022.05.035 http://dx.doi.org/10.3969/j.issn.1006-2084.2022.05.035 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621