PET/CT影像组学鉴别肿瘤标志物阴性且FDG摄取阳性孤立性肺结节的临床价值

Clinical value of PET/CT radiomics in differentiating tumor marker-negative, FDG-avid solitary pulmonary nodules

摘要

目的 探讨¹⁸F-氟代脱氧葡萄糖（¹⁸F-FDG）正电子发射断层扫描/计算机断层扫描（PET/CT）影像组学模型在鉴别肿瘤标志物阴性、FDG摄取阳性孤立性肺结节良恶性中的临床价值。方法 收集2019年1月至2024年12月保定市第一中心医院收治的130例此类肺结节患者资料，按7∶3比例随机分为训练集（ n = 91）与内部测试集（ n = 39）。另从TCIA公共数据库中收集符合条件的独立病例作为外部验证集（ n = 45）。使用LIFEx软件提取PET及CT影像组学特征，经组内相关系数（ICC）稳定性评估、Spearman相关性分析及LASSO回归筛选核心特征。分别构建逻辑回归（LR）、随机森林（RF）和支持向量机（SVM）模型，采用受试者操作特征（ROC）曲线与决策曲线分析（DCA）评估并比较各模型性能。结果 最终筛选出8个关键影像组学特征（PET与CT各4个）。在内部测试集中，LR模型效能最优，且曲线下面积（AUC）为0.819（95%CI为0.738 ～ 0.892），灵敏度与特异度分别为0.792与0.800。在外部验证集中，LR模型仍保持最佳鉴别性能，AUC为0.801（95%CI为0.713 ～ 0.905），优于RF模型（AUC = 0.721， P = 0.041）与SVM模型（AUC = 0.717， P = 0.028）。DCA显示，LR模型在内部测试集及外部验证集中均能提供较高的临床净获益。结论 针对肿瘤标志物阴性且FDG摄取阳性的难辨性孤立性肺结节，PET/CT影像组学模型（尤其是LR模型）具有良好的鉴别效能与泛化能力，可为术前决策提供客观的量化参考。

Abstract

Objective To investigate the clinical value of an ¹⁸F-FDG PET/CT radiomics model in differentiating benign from malignant solitary pulmonary nodules with negative tumor markers and positive FDG uptake. Methods The data of 130 patients with such pulmonary nodules admitted to Baoding No. 1 Central Hospital from January 2019 to December 2024 were collected. The patients were randomly divided into a training set ( n = 91) and an internal test set ( n = 39) at a ratio of 7:3. Eligible independent cases were additionally collected from the Cancer Imaging Archive (TCIA) public database as an external validation set ( n = 45). PET and CT radiomics features were extracted using LIFEx software. Core features were selected through intraclass correlation coefficient (ICC) stability assessment, Spearman correlation analysis, and least absolute shrinkage and selection operator (LASSO) regression. Logistic regression (LR), random forest (RF), and support vector machine (SVM) models were constructed separately. Receiver operating characteristic (ROC) curves and decision curve analysis (DCA) were used to evaluate and compare the performance of the models. Results A total of 8 key radiomics features were ultimately selected, including 4 PET features and 4 CT features. In the internal test set, the LR model showed the best performance, with an area under the curve (AUC) of 0.819 (95% CI: 0.738-0.892), a sensitivity of 0.792, and a specificity of 0.800. In the external validation set, the LR model also maintained the best discriminative performance, with an AUC of 0.801 (95% CI: 0.713-0.905), which was significantly higher than those of the RF model (AUC = 0.721, P = 0.041) and the SVM model (AUC = 0.717, P = 0.028). DCA demonstrated that the LR model yielded greater clinical net benefit in both the internal test set and the external validation set. Conclusions For difficult-to-differentiate solitary pulmonary nodules with negative tumor markers and high FDG uptake, PET/CT radiomics models, particularly the LR model, have good discriminative performance and generalizability and can provide an objective quantitative reference for preoperative decision-making.

关键词

肺结节 / 正电子发射计算机断层显像 / 影像组学 / 预测模型 / 肿瘤标志物阴性

Key words

Solitary pulmonary nodule / PET/CT / Radiomics / Prediction model / Negative tumor markers

参考文献

[1] 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.

[2] 中华医学会呼吸病学分会, 中国肺癌防治联盟专家组. 肺结节诊治中国专家共识（2024年版）[J]. 中华结核和呼吸杂志, 2024, 47(8): 716-729. DOI: 10.3760/cma.j.cn112147-20231109-00300.

[3] Chinese Thoracic Society, Chinese Medical Association; Chinese Alliance Against Lung Cancer Expert Group. Chinese expert consensus on diagnosis and treatment of pulmonary nodules (2024)[J]. Chin J Tuberc Respir Dis, 2024, 47(8): 716-729. DOI: 10.3760/cma.j.cn112147-20231109-00300.

[4] 武强, 张博友, 孙飞, 等 . 多发性肺结节影像学管理现状及进展[J]. 中国胸心血管外科临床杂志, 2021, 28(3): 363-367. DOI: 10.7507/1007-4848.202005083.

[5] Wu Q, Zhang B Y, Sun F, et al. Current status and advancement of medical imaging management for multiple pulmonary nodules[J]. Chin J Clin Thorac Cardiovasc Surg, 2021, 28(3): 363-367. DOI: 10.7507/1007-4848.202005083.

[6] 赫捷, 李霓, 陈万青, 等 . 中国肺癌筛查与早诊早治指南（2021, 北京）[J]. 中华肿瘤杂志, 2021, 43(3): 243-268. DOI: 10.3760/cma.j.cn112152-20210119-00060.

[7] He J, Li N, Chen W Q, et al. China guideline for the screening and early detection of lung cancer (2021, Beijing)[J]. Chin J Oncol, 2021, 43(3): 243-268. DOI: 10.3760/cma.j.cn112152-20210119-00060.

[8] 刘永, 陈鹏, 孙振国, 等 . ¹⁸F-FDG PET/CT诊断孤立性肺结节良恶性的效能分析 [J]. 中华核医学与分子影像杂志, 2020, 40(7): 411-414. DOI: 10.3760/cma.j.cn321828-20191125-00182.

[9] Liu Y, Chen P, Sun Z G, et al. Efficacy of¹⁸F-FDG PET/CT in the differential diagnosis of benign and malignant solitary pulmonary nodules [J]. Chin J Nucl Med Mol Imaging, 2020, 40(7): 411-414. DOI: 10.3760/cma.j.cn321828-20191125-00182.

[10] Corica F, de Feo M S, Stazza M L, et al. Qualitative and semiquantitative parameters of¹⁸F-FDG-PET/CT as predictors of malignancy in patients with solitary pulmonary nodule [J]. Cancers, 2023, 15(4): 1000. DOI: 10.3390/cancers15041000.

[11] Koo J M, Kim J, Lee J, et al. Deciphering the intratumoral histologic heterogeneity of lung adenocarcinoma using radiomics[J]. Eur Radiol, 2025, 35(8): 4861-4872. DOI: 10.1007/s00330-025-11397-4.

[12] 冯源, 兰晓莉. 影像组学介绍[J]. 中华核医学与分子影像杂志, 2023, 43(1): 55-60. DOI: 10.3760/cma.j.cn321828-20211130-00427.

[13] Feng Y, Lan X L. Introduction to radiomics[J]. Chin J Nucl Med Mol Imaging, 2023, 43(1): 55-60. DOI: 10.3760/cma.j.cn321828-20211130-00427.

[14] 张伟娜, 钟李长, 师琳, 等 . 超声影像组学特征联合miRNA-34a表达水平对乳腺癌新辅助化疗患者病理完全缓解的预测能力[J]. 新医学, 2025, 56(7): 645-653. DOI: 10.12464/j.issn.0253-9802.2025-0023.

[15] Zhang W N, Zhong L C, Shi L, et al. Predictive ability of ultrasound radiomics features combined with miRNA-34a expression levels for pathological complete response in breast cancer patients receiving neoadjuvant chemotherapy[J]. J New Med, 2025, 56(7): 645-653. DOI: 10.12464/j.issn.0253-9802.2025-0023.

[16] Qin L, Zhao L, Li X M, et al. A CT-based radiomics nomogram incorporating adipose tissue to differentiate invasive adenocarcinomas among part-solid pulmonary nodules[J]. BMC Cancer, 2025, 25(1): 1471. DOI: 10.1186/s12885-025-14875-6.

[17] Miao S, Xuan Q, Xie H, et al. An integrated nomogram combining deep learning and radiomics for predicting malignancy of pulmonary nodules using CT-derived nodules and adipose tissue: a multicenter study[J]. Cancer Med, 2024, 13(21): e70372. DOI: 10.1002/cam4.70372.

[18] Jha A K, Sherkhane U B, Purandare N C, et al. Positron emission tomography imaging biomarker and artificial intelligence for the characterization of solitary pulmonary nodule[J]. Front Nucl Med, 2025, 5: 1611823. DOI: 10.3389/fnume.2025.1611823.

[19] Boellaard R, Delgado-Bolton R, Oyen W J G, et al. FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0[J]. Eur J Nucl Med Mol Imaging, 2015, 42(2): 328-354. DOI: 10.1007/s00259-014-2961-x.

[20] Nioche C, Orlhac F, Boughdad S, et al. LIFEx: a freeware for radiomic feature calculation in multimodality imaging to accelerate advances in the characterization of tumor heterogeneity[J]. Cancer Res, 2018, 78(16): 4786-4789. DOI: 10.1158/0008-5472.CAN-18-0125.

[21] Tibdewal A, Patil M, Misra S, et al. Optimal standardized uptake value threshold for auto contouring of gross tumor volume using positron emission tomography/computed tomography in patients with operable nonsmall-cell lung cancer: comparison with pathological tumor size[J]. Indian J Nucl Med, 2021, 36(1): 7-13. DOI: 10.4103/ijnm.IJNM_134_20.

[22] Zwanenburg A, Vallières M, Abdalah M A, et al. The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping[J]. Radiology, 2020, 295(2): 328-338. DOI: 10.1148/radiol.2020191145.

[23] Chen W, Liu Q, Li H, et al. The prediction performance analysis of the Lasso model with convex non-convex sparse regularization[J]. Algorithms, 2025, 18(4): 195. DOI: 10.3390/a18040195.

[24] Zhang Y, Huang W, Jiao H, et al. PET radiomics in lung cancer: advances and translational challenges[J]. EJNMMI Phys, 2024, 11(1): 81. DOI: 10.1186/s40658-024-00685-5.

[25] Kapsoritakis N, Tsitoura F, Stathaki M, et al. Clinical impact of FDG PET/CT in pulmonary nodule characterization: current perspectives on dual-time-point imaging and semi-quantitative imaging metrics[J]. Cancers, 2025, 17(20): 3353. DOI: 10.3390/cancers17203353.

[26] Raad R A, Garrana S, Moreira A L, et al. Imaging and management of subsolid lung nodules[J]. Radiologic Clin N Am, 2025, 63(4): 517-535. DOI: 10.1016/j.rcl.2024.12.002.

[27] Du Y, Yuan C, Ni X, et al. Overdiagnosis in low-dose CT lung cancer screening: a systematic review and meta-analysis of overall magnitude and subgroup variations[J]. Int J Cancer, 2025, 157(11): 2269-2282. DOI: 10.1002/ijc.70049.

[28] 李红英, 胡鑫, 宋瑞祥. 高分辨CT影像学特征对孤立性肺结节良恶性的鉴别诊断效能[J]. 海南医学, 2022, 33(19): 2540-2543. DOI: 10.3969/j.issn.1003-6350.2022.19.025.

[29] Li H Y, Hu X, Song R X. Analysis of high resolution CT imaging features in the differential diagnosis of solitary pulmonary nodules[J]. Hainan Med J, 2022, 33(19): 2540-2543. DOI: 10.3969/j.issn.1003-6350.2022.19.025.

[30] 沈天赐, 宋鑫洋, 胡翔宇, 等 . 基于薄层增强CT图像探究影像组学在肺结节良恶性诊断价值[J]. 中国CT和MRI杂志, 2025, 23(4): 75-78. DOI: 10.3969/j.issn.1672-5131.2025.04.022.

[31] Shen T C, Song X Y, Hu X Y, et al. Exploring the value of radiomics in the diagnosis of benign and malignant pulmonary nodules based on thin-section enhanced CT images[J]. Chin J CT MRI, 2025, 23(4): 75-78. DOI: 10.3969/j.issn.1672-5131.2025.04.022.

[32] Palumbo B, Bianconi F, Palumbo I, et al. Value of shape and texture features from¹⁸F-FDG PET/CT to discriminate between benign and malignant solitary pulmonary nodules: an experimental evaluation [J]. Diagnostics (Basel), 2020, 10(9): 696. DOI: 10.3390/diagnostics10090696.

[33] Agüloğlu N, Aksu A, Unat D S . Machine learning approach using 18F-FDG PET-based radiomics in differentiation of lung adenocarcinoma with bronchoalveolar distribution and infection[J]. Nucl Med Commun, 2023, 44(4): 302-308. DOI: 10.1097/mnm.0000000000001667.

[34] Zheng J, Hao Y, Guo Y, et al. An 18F-FDG-PET/CT-based radiomics signature for estimating malignancy probability of solitary pulmonary nodule[J]. Clin Respir J, 2024, 18(5): e13751. DOI: 10.1111/crj.13751.

[35] Zhang L, Wan X, Ji K, et al. The study of dual-phase¹⁸F-FDG PET/CT-based models in predicting malignant solitary pulmonary lesions [J]. Sci Rep, 2025, 15(1): 34896. DOI: 10.1038/s41598-025-18677-5.

[36] Collins G S, Moons K G M, Dhiman P, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods[J]. BMJ, 2024, 385: e078378. DOI: 10.1136/bmj-2023-078378.

[37] 王贺真, 边芳, 童玉洁, 等 . 基于MRI影像组学联合临床特征的机器学习模型预测宫颈鳞癌组织学分级的价值[J]. 新医学, 2024, 55(3): 176-183. DOI: 10.3969/j.issn.0253-9802.2024.03.005.

[38] Wang H Z, Bian F, Tong Y J, et al. Value of machine learning model based on MRI radiomics in predicting histological grade of cervical squamous cell carcinoma[J]. J New Med, 2024, 55(3): 176-183. DOI: 10.3969/j.issn.0253-9802.2024.03.005.

[39] Vishraj R, Gupta S, Singh S . Evaluation of feature selection methods utilizing random forest and logistic regression for lung tissue categorization using HRCT images[J]. Expert Syst, 2023, 40(8): e13320. DOI: 10.1111/exsy.13320.

[40] Yao Z, Wang Y, Wu Y, et al. Leveraging machine learning with dynamic 18F-FDG PET/CT: integrating metabolic and flow features for lung cancer differential diagnosis[J]. Eur J Nucl Med Mol Imaging, 2025, 52(10): 3807-3819. DOI: 10.1007/s00259-025-07231-0.

[41] Fridgeirsson E A, Williams R, Rijnbeek P, et al. Comparing penalization methods for linear models on large observational health data[J]. J Am Med Inform Assoc, 2024, 31(7): 1514-1521. DOI: 10.1093/jamia/ocae109.

基金资助

河北省医学科学研究课题计划项目(20232029)

保定市科技计划项目(2241ZF262)