基于XGBoost人工智能结合CT构建甲状腺癌颈部淋巴结转移预测模型

doi:10.6040/j.issn.1673-3770.1.2020.031

摘要/Abstract

摘要： 目的基于极端梯度提升算法人工智能建立甲状腺癌患者术前CT颈部淋巴结转移的预测模型,为临床制定规范的治疗方案提供参考依据。方法回顾性分析2017年10月~2019年5月38例甲状腺乳头状癌患者临床资料,共纳入135个淋巴结数据。采集甲状腺癌淋巴结转移相关变量及CT参数,基于XGBoost人工智能评估数据特征属性重要性,建立淋巴结转移预测模型,采用五折交叉验证方法训练测试模型。结果基于XGBoost人工智能甲状腺癌CT淋巴结转移预测模型准确率平均为87.41%,优于支持向量机机器学习算法模型79.22%。淋巴结强化、强化不均匀、原发灶同侧淋巴结及淋巴结有包膜侵犯是提示淋巴结转移的重要的CT特征属性。结论基于XGBoost人工智能建立的甲状腺癌患者术前CT颈部淋巴结转移的预测模型准确率高,可以帮助临床医师术前判断甲状腺癌是否伴有颈部淋巴结转移,评估肿瘤TNM分期,并制定规范的手术治疗方案。

关键词: 甲状腺癌, 淋巴结转移, XGBoost, 人工智能, 预测模型

Abstract: Objective Incorporating eXtreme Gradient Boosting(XGBoost)artificial intelligence, we aimed to build a predictive model using pre-operative enhanced computed tomography(CT)of cervical lymph node metastasis in patients with thyroid cancer, to provide a reference for pre-operative planning. Methods The clinical data of 38 patients with thyroid papillary carcinoma from October 2017 to May 2019 were retrospectively analyzed. A total of 135 lymph nodes were included. Using XGBoost artificial intelligence, the lymph node metastasis prediction model was established, and the accuracy of the prediction model was tested. Results The average accuracy of the XGBoost model was 87.41%, which was higher than that of the SVM model(79.2%). Important CT characteristics that are indicative of lymph node metastasis include degree and distribution of enhancement, location, and capsule invasion. Conclusion The predictive model of cervical lymph node metastasis in patients with thyroid cancer exhibits high accuracy and could help in the pre-operative evaluation of cervical lymph node metastasis, tumor staging, and surgical procedures.

Key words: Thyroid cancer, Lymph node metastasis, XGBoost, Artificial intelligence, Prediction model

中图分类号:

R736

陈海兵, 卫亚楠, 许晓泉, 陈曦. 基于XGBoost人工智能结合CT构建甲状腺癌颈部淋巴结转移预测模型[J]. 山东大学耳鼻喉眼学报, 2020, 34(3): 40-45.

CHEN Haibing, WEI Ya'nan, XU Xiaoquan, CHEN Xi. Prediction of cervical lymph node metastasis in papillary thyroid cancer based on XGBoost artificial intelligence and enhanced computed tomography[J]. Journal of Otolaryngology and Ophthalmology of Shandong University, 2020, 34(3): 40-45.

参考文献

[1] Rosenbaum MA, McHenry CR. Contemporary management of papillary carcinoma of the thyroid gland[J]. Expert Rev Anticancer Ther, 2009, 9(3): 317-329. doi:10.1586/14737140.9.3.317.
[2] Hay ID. Papillary thyroid carcinoma[J]. medicina clínica, 1990, 132(8):675-676.
[3] Leboulleux S, Rubino C, Baudin E, et al. Prognostic factors for persistent or recurrent disease of papillary thyroid carcinoma with neck lymph node metastases and/or tumor extension beyond the thyroid capsule at initial diagnosis[J]. J Clin Endocrinol Metab, 2005, 90(10): 5723-5729. doi:10.1210/jc.2005-0285.
[4] 王莎莎, 林岩松, 梁军, 等. 甲状腺乳头状癌首次治疗术式对预后影响的分析[J]. 中华肿瘤防治杂志, 2012, 19(10): 761-765. doi:10.16073/j.cnki.cjcpt. 2012.10.014. WANG Shasha, LIN Yansong, LIANG Jun, et al. Initial surgical procedure impact on prognosis of papillary thyroid carcinoma[J]. Chinese Journal of Cancer Prevention and Treatment, 2012, 19(10): 761-765. doi:10.16073/j.cnki.cjcpt.2012.10.014.
[5] 高晓倩, 姜震, 耿琛琛, 等. 术前超声评估分化型甲状腺癌颈部淋巴结转移[J]. 山东大学耳鼻喉眼学报, 2019, 33(1): 135-139. doi: 10.6040/j.issn.1673-3770.0.2018.443. GAO Xiaoqian, JIANG Zhen, GENG Chenchen, et al. Preoperative ultrasonography in detecting cervical lymph node metastasis in differentiated thyroid cancer[J]. Journal of Otolaryngology and Ophthalmology of Shandong University, 2019, 33(1): 135-139.
[6] 施燕芸, 李念芬, 孙红光, 等. 超声造影预测甲状腺乳头状癌颈部淋巴结转移的临床价值[J]. 临床超声医学杂志, 2018, 20(8): 526-530. doi:10.16245/j.cnki.issn1008-6978.2018.08.007. SHI Yanyun, LI Nianfen, SUN Hongguang, et al. Clinical value of contrast-enhanced ultrasonography in predicting cervical lymph node metastasis of papillary thyroid carcinoma[J]. Journal of Clinical Ultrasound in Medicine, 2018, 20(8): 526-530. doi:10.16245/j.cnki.issn1008-6978.2018.08.007.
[7] 樊金芳, 余小情, 陶玲玲, 等. 超声弹性成像及超声造影预测甲状腺乳头状癌淋巴结转移的价值探讨[J]. 中国医学计算机成像杂志, 2019, 25(1): 67-71. doi:10.19627/j.cnki.cn31-1700/th.2019.01.014. FAN Jinfang, YU Xiaoqing, TAO Lingling, et al. Value of ultrasound elastography and contrast ultrasoundin predicting lymph node metastasis of thyroid papillary carcinoma[J]. Chinese Computed Medical Imaging, 2019, 25(1): 67-71. doi:10.19627/j.cnki.cn31-1700/th.2019.01.014.
[8] 林启强, 韩志江, 舒艳艳, 等. CT在评估甲状腺乳头状癌中央组淋巴结转移中的价值[J]. 中国临床医学影像杂志, 2015, 26(3): 162-165, 205. LIN Qiqiang, HAN Zhijiang, SHU Yanyan, et al. Value of CT in diagnosing central lymph node metastasis of papillary thyroid carcinoma[J]. Journal of China Clinic Medical Imaging, 2015, 26(3): 162-165, 205.
[9] Chen TQ, Guestrin C. XGBoost[C] //Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco California USA. New York, NY, USA: ACM, 2016. doi:10.1145/2939672.2939785.
[10] Koo BS, Choi EC, Yoon YH, et al. Predictive factors for ipsilateral or contralateral central lymph node metastasis in unilateral papillary thyroid carcinoma[J]. Ann Surg, 2009, 249(5): 840-844. doi:10.1097/sla.0b013e3181a40919.
[11] Ito Y, Jikuzono T, Higashiyama T, et al. Clinical significance of lymph node metastasis of thyroid papillary carcinoma located in one lobe[J]. World J Surg, 2006, 30(10): 1821-1828. doi:10.1007/s00268-006-0211-5.
[12] 张正华, 黄建强, 韩丹, 等. 甲状腺癌的CT表现与颈淋巴结转移的相关性[J]. 中国医学影像学杂志, 2013, 21(11): 804-807, 811. doi: 10.3969/j.issn.1005-5185.2013.11.002. ZHANG Zhenghua, HUANG Jianqiang, HAN Dan, et al. CT manifestations of thyroid carcinoma and its correlation with lymph node metastasis[J]. Chinese Journal of Medical Imaging, 2013, 21(11): 804-807, 811. doi: 10.3969/j.issn.1005-5185.2013.11.002.
[13] 王天笑, 宋韫韬, 徐国辉, 等. 细针穿刺技术在甲状腺乳头状癌侧颈淋巴转移中的预测价值[J]. 中华耳鼻咽喉头颈外科杂志, 2019, 54(1): 23-27. doi: 10.3760/cma.j.issn.1673-0860.2019.01.006. WANG Tianxiao, SONG Yuntao, XU Guohui, et al. Fine-needle aspiration for the diagnosis of lymph node metastasis in papillary thyroid carcinoma[J]. Chinese Journal of Otorhinolaryngology Head and Neck Surgery, 2019, 54(1): 23-27. doi: 10.3760/cma.j.issn.1673-0860.2019.01.006.
[14] 葛智成, 屈翔, 滕长胜, 等. USG和增强CT对甲状腺癌与中央区淋巴结转移的研究[J]. 国际外科学杂志, 2012(2): 87-90. doi: 10.3760/cma.j.issn.1673-4203.2012.02.005. GE Zhicheng, QU Xiang, TENG Changsheng, et al. Study of ultrasonography and contrast-enhanced computer tomography in the diagnosis of thyroid carcinoma and lymph node metastasis[J]. International Journal of Surgery, 2012(2): 87-90. doi: 10.3760/cma.j.issn.1673-4203.2012.02.005.
[15] 李强, 赵博文, 吕江红, 等. FNA-Tg测定在细针穿刺诊断甲状腺癌术后侧颈区可疑肿大淋巴结中的应用价值[J]. 中华耳鼻咽喉头颈外科杂志, 2016, 51(5):378-382. doi: 10.3760/cma.j.issn.1673-0860.2016.05.012. LI Qiang, ZHAO Bowen, LV Jianghong, et al. The value of thyroglobulin measurement in fine-needle aspiration for diagnosis of suspicious lymph nodes in patients with thyroid carcinoma after thyroidectomy[J]. Chinese JOurnal of Otorhinolaryngology Head and Neck Surgery, 2016, 51(5):378-382. doi: 10.3760/cma.j.issn.1673-0860.2016.05.012.
[16] Seo SH, Lee JH, Soh EY. In thyroid cancer patients, is preoperative FNAB-C reliable for prediction of lateral cervical LN metastasis?[J]. Korean J Endocr Surg, 2014, 14(2): 76. doi:10.16956/kjes.2014.14.2.76.
[17] 纪宇楠. 基于随机森林构建滤泡型甲状腺癌远处转移预测模型[D]. 沈阳: 中国医科大学, 2018.
[18] 郭芳琪, 赵佳琦, 刘晟. 人工智能自动检测系统在甲状腺结节术前超声诊断中的应用[J]. 第二军医大学学报, 2019, 40(11): 1183-1189. doi:10.16781/j.0258-879x.2019.11.1183. GUO Fangqi, ZHAO Jiaqi, LIU Sheng. Application of artificial intelligence automatic detection system in preoperative ultrasonic diagnosis of thyroid nodules[J]. Academic Journal of Second Military Medical University, 2019, 40(11): 1183-1189. doi:10.16781/j.0258-879x.2019.11.1183.
[19] 李婷婷, 卢漫, 巫明钢, 等. 计算机辅助诊断系统对甲状腺结节的诊断价值研究[J]. 中华医学超声杂志(电子版), 2019, 16(9): 660-664. doi: 10.3877/cma.j.issn.1672-6448.2019.09.004. LI Tingting, LU Man, WU Minggang, et al. Performance of computer-aided diagnosis system versus radiologists in diagnosis of thyroid nodules[J]. Chinese Journal of Medical Ultrasound(Electronic Edition), 2019, 16(9): 660-664. doi: 10.3877/cma.j.issn.1672-6448.2019.09.004.
[20] Torlay L, Perrone-Bertolotti M, Thomas E, et al. Machine learning-XGBoost analysis of language networks to classify patients with epilepsy[J]. Brain Inform, 2017, 4(3): 159-169. doi:10.1007/s40708-017-0065-7.
[21] Zhong JC, Sun YS, Peng W, et al. XGBFEMF: an XGBoost-based framework for essential protein prediction[J]. IEEE Trans Nanobioscience, 2018, 17(3): 243-250. doi:10.1109/TNB.2018.2842219.
[22] Li W, Yin YB, Quan XW, et al. Gene expression value prediction based on XGBoost algorithm[J]. Front Genet, 2019, 10: 1077. doi:10.3389/fgene.2019.01077.
[23] Pang L, Wang JJ, Zhao LL, et al. A novel protein subcellular localization method with CNN-XGBoost model for Alzheimer's disease[J]. Front Genet, 2018, 9: 751. doi:10.3389/fgene.2018.00751.
[24] Zheng HT, Yuan JB, Chen L. Short-term load forecasting using EMD-LSTM neural networks with a xgboost algorithm for feature importance evaluation[J]. Energies, 2017, 10(8): 1168. doi:10.3390/en10081168.
[25] Li CB, Zheng XS, Yang ZK, et al. Predicting short-term electricity demand by combining the advantages of Arma and XGBoost in fog computing environment[J]. Wirel Commun Mob Comput, 2018, 2018: 1-18. doi:10.1155/2018/5018053.
[26] Jeong JJ, Lee YS, Lee SC, et al. A scoring system for prediction of lateral neck node metastasis from papillary thyroid cancer[J]. J Korean Med Sci, 2011, 26(8): 996-1000. doi:10.3346/jkms.2011.26.8.996.
[27] Yang YL, Chen CZ, Chen ZM, et al. Prediction of central compartment lymph node metastasis in papillary thyroid microcarcinoma[J]. Clin Endocrinol(Oxf), 2014, 81(2): 282-288. doi:10.1111/cen.12417.

相关文章 15

[1]	杨冠英,李元彬. 人工智能在干眼管理中的应用进展[J]. 山东大学耳鼻喉眼学报, 2026, 40(3): 115-120.
[2]	朱明琼,李征,刘茹,田涛,彭婧利,吕倩怡,谭华霞. 基于OCT/OCTA的AI筛查系统在抗VEGF治疗糖尿病性黄斑水肿患者效果评价中的应用[J]. 山东大学耳鼻喉眼学报, 2026, 40(1): 68-73.
[3]	程卓, 梁辉, 邢鲁民. 深度学习技术在咽喉内镜应用中的研究进展及前景分析[J]. 山东大学耳鼻喉眼学报, 2026, 40(1): 112-119.
[4]	张中华,王辉,王俊鑫,张美玲,张建,张恩东,潘新良. 腋窝入路腔镜手术和传统颈部开放手术治疗T1甲状腺癌的安全性和有效性对比研究[J]. 山东大学耳鼻喉眼学报, 2025, 39(4): 161-167.
[5]	吕勇,冯云. 双感官障碍人群心血管疾病风险的预测模型:基于CHARLS的分析[J]. 山东大学耳鼻喉眼学报, 2025, 39(3): 122-128.
[6]	杨鸣,刘雪霞,张华. m6A识别蛋白IGF2BPs家族在头颈肿瘤中的研究进展[J]. 山东大学耳鼻喉眼学报, 2025, 39(3): 153-161.
[7]	张国明,魏文斌,林浩添,迟玮,张少冲,赵培泉,雷柏英,陈有信,王雨生,何明光,梁建宏,卢海,陆方,黄欣,梁小玲,赵欣予,吴桢泉,余震,崔凯璇,刘亚玲,项道满,陈长征,张自峰,林铎儒,于珊珊,孙悦,檀韬,陈燕先,彭婕,董力,程湧,朱雪梅,杨鹏,陈少滨. 人工智能技术辅助早产儿视网膜病变诊疗专家共识(2025)[J]. 山东大学耳鼻喉眼学报, 2025, 39(2): 1-5.
[8]	李培培,卢彦青,侯楠. 机器学习预测模型在突发性聋中的临床应用研究[J]. 山东大学耳鼻喉眼学报, 2025, 39(2): 145-151.
[9]	沈嘉琪,李潇飒,毕燕龙,张敬法. 人工智能在DME筛查、诊断和预后中的应用[J]. 山东大学耳鼻喉眼学报, 2024, 38(5): 153-159.
[10]	张琳,辛运超,尚小领,解琪,刘亚超. cN0期甲状腺微小乳头状癌颈部中央区及侧颈区淋巴结转移率、危险因素及清扫策略的临床研究[J]. 山东大学耳鼻喉眼学报, 2024, 38(4): 62-69.
[11]	王靖淞,余灿,张西,邓启成,谢卓良,赵锐,温蓓,刘海. KRT4在喉癌及癌旁正常黏膜组织中的表达及临床意义[J]. 山东大学耳鼻喉眼学报, 2024, 38(3): 12-17.
[12]	谢玉林,雷大鹏. 人工智能在头颈部鳞状细胞癌淋巴结转移的病理研究进展[J]. 山东大学耳鼻喉眼学报, 2024, 38(3): 124-129.
[13]	吴丽丽,曲毅. OCTA在病理性近视脉络膜新生血管应用及其在人工智能的研究进展[J]. 山东大学耳鼻喉眼学报, 2024, 38(2): 144-149.
[14]	张一彤,李青香,石争浩,尚磊,袁钰淇,曹子讷,麻莉娜,刘海琴,任晓勇,施叶雯. 阻塞性睡眠呼吸暂停儿童睡眠结构研究及睡眠结构判读模型建立[J]. 山东大学耳鼻喉眼学报, 2023, 37(6): 126-132.
[15]	杜曰山一,王鲜,张国明. 人工智能辅助早产儿视网膜病变诊疗新进展[J]. 山东大学耳鼻喉眼学报, 2023, 37(3): 157-162.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed