切换至 "中华医学电子期刊资源库"

中华临床医师杂志(电子版) ›› 2024, Vol. 18 ›› Issue (08) : 787 -792. doi: 10.3877/cma.j.issn.1674-0785.2024.08.014

综述

人工智能应用于多发性肺结节诊断的研究进展
孙铭远1, 褚恒2, 徐海滨3, 张哲2,()   
  1. 1.261000 山东潍坊,山东第二医科大学临床医学院
    2.266011 山东青岛,青岛市市立医院胸外科
    3.266011 山东青岛,青岛市市立医院影像科
  • 收稿日期:2024-06-27 出版日期:2024-08-15
  • 通信作者: 张哲
  • 基金资助:
    国家自然科学基金项目(22204152)

Progress in application of artificial intelligence in diagnosis of multiple pulmonary nodules

Mingyuan Sun1, Heng Chu2, Haibin Xu3, Zhe Zhang2,()   

  1. 1.School of Clinical Medicine,Shandong Second Medical University,Weifang 261000,China
    2.Department of Thoracic surgery,Qingdao Municipal Hospital,Qingdao 266011,China
    3.Department of Imaging,Qingdao Municipal Hospital,Qingdao 266011,China
  • Received:2024-06-27 Published:2024-08-15
  • Corresponding author: Zhe Zhang
引用本文:

孙铭远, 褚恒, 徐海滨, 张哲. 人工智能应用于多发性肺结节诊断的研究进展[J]. 中华临床医师杂志(电子版), 2024, 18(08): 787-792.

Mingyuan Sun, Heng Chu, Haibin Xu, Zhe Zhang. Progress in application of artificial intelligence in diagnosis of multiple pulmonary nodules[J]. Chinese Journal of Clinicians(Electronic Edition), 2024, 18(08): 787-792.

肺癌,作为世界范围内发病率和死亡率最高的恶性肿瘤之一,早期影像学表现为肺结节。其中,多发性肺结节因其逐年增高的检出率和特殊性而受到广泛关注。因此对肺结节性质进行正确预测是肺癌早期诊治的关键。近年来,人工智能(artificial intelligence,AI)技术与医学的结合在肺结节的诊断方面有了很大的进展,特别是通过深度学习、机器学习、放射组学等技术对肺结节的特征和性质进行分析及预测。这些方法极大地提高了肺癌早期筛查的效率和准确性,对肺癌的诊治具有重要的指导作用。本文综述了AI 应用于肺结节诊断的进展,尤其是多发性肺结节,并分析了AI 技术目前的优势和不足以及未来的发展方向,以期为多发性肺结节的诊治提供新思路、新方法。

Lung cancer is a malignancy with the highest incidence and mortality rates worldwide,and the early-stage imaging findings of lung cancer are often pulmonary nodules.Multiple pulmonary nodules have attracted great attention due to their increasing detection rate and high specificity.Thus,accurate prediction of the nature of pulmonary nodules is crucial for early diagnosis and treatment of lung cancer.In recent years, there have been great improvements in the diagnosis of pulmonary nodules due to the integration of artificial intelligence (AI) into medical technologies, particularly the application of deep learning, machine learning, and radiomics to analyze and predict the features and nature of pulmonary nodules.These approaches have dramatically enhanced the efficiency and accuracy of early screening for lung cancer, offering important guidance for the diagnosis and treatment of this malignancy.This article reviews the progress in AI application in the diagnosis of pulmonary nodules, especially in multiple pulmonary nodules, and analyzes the current advantages and limitations of AI technology as well as future development directions, aiming to provide new ideas and methods for the diagnosis and treatment of multiple pulmonary nodules.

表1 人工智能应用于肺结节诊断相关研究
作者 年份 结节类型 AI工具 目的 样本 结果
Qiao Liu等[32] 2024 - 机器学习 建立基于机器学习的新型多分类预测模型来预测肺结节恶性肿瘤的概率 914名患者 RF模型对ML、PL和BL模型的AUC值分别为0.80、0.90和0.75
Yuhei Takeshita等[33] 2024 - U-net模型 验证人工智能对肺结节进行自动跟踪的可行性 40例患者共49个肺结节 随访CT检查368次,每次评估的成功率为94%,逐个结节评估的成功率为78%
Wei Fan等[34] 2024 - 深度学习 人工智能CT密度检测和处理肺良恶性结节中的诊断价值 130名患者226个肺结节 人工智能识别肺结节的敏感性为94.69%,识别高危结节的敏感性为95.40%,特异性为75%。
Haochuan Zhang等[35] 2023 亚厘米亚实性结节 LR&SVM&XGBoost 预测亚厘米亚实性肺结节的组织学侵袭性 177名患者,203个结节 LR训练集AUC=0.743(0.6610.824);测试集AUC=0.803(0.694~0.913)
SVM训练集AUC=0.828(0.760.896);测试集AUC=0.726(0.598~0.854)
XGBoost训练集AUC=0.917(0.8690.965);测试集AUC=0.874(0.776~0.972)
Junhao Mu等[36] 2023 实性结节 深度学习 预测肺实性结节的恶性和转移 1571例 预测恶性肿瘤AUC=80.37%
预测转移AUC=86.44%
T.-W.Tang等[37] 2023 - NoduleNet深度学习模型 结合深度学习模型建立基于lung-rads标准的肺结节诊断流程 1186个结节 总体诊断准确率为0.75
Lichuan Zhang等[29] 2023 单发 深度学习 探讨人工智能辅助诊断系统在肺结节预测中的价值 260例 人工智能预测准确率为77.57%,敏感性为89.60%,特异性为48.28%。
Wei Pan等[6] 2023 - 人工智能 探讨人工智能在CT扫描中诊断肺结节的准确性 309名患者的360个结节 人工智能诊断肺结节的准确率为81.94%,漏诊率为15.14%,误诊率为24.77%,真阴性率为75.23%
Yanqing Ma等[27] 2022 单发和多发 CT delta-放射组学 评估多发性原发性肺腺癌(MPLA)和孤立性原发性肺腺癌(SPLA)之间的差异 1094例患者,含268例MPLA和826例SPLA A组随访间隔3~12个月,105例MPLA vs 145例SPLA)训练集AUC为0.972(95%CI:0.951~0.989),测试集AUC为0.798(95%CI:0.704~0.892)
B组(随访间隔13~24个月,68例MPLA vs. 96例SPLA)训练集AUC为0.989(95%CI:0.978~0.997),测试集AUC为0.821(95%CI:0.708~0.915)
C组(随访间隔25~48个月,52例MPLA vs 79例SPLA)训练集AUC为0.998(95%CI:0.993~1.000),测试集AUC为0.853(95%CI:0.752~0.940)
Zhang Kai等[38] 2022 多发 PKUML模型 评估模型判定MPN性质的效力 287名患者,446个实结节 AUC=0.883(95%CI:0.849~0.917),敏感度为74.2%,特异度为88.6%(阈值0.541)Cindex=0.883
Rui Zhang等[39] 2022 实性结节 CNN&RF 评估放射组学和深度学习技术对实性肺结节的诊断 720个实性肺结节 CNN模型AUC=0.819(95%CI:0.760~0.877),敏感性0.778,特异性0.788,准确性0.783,与放射组学模型差异无统计学意义
Kezhong Chen等[28] 2021 多发 PKU-M模型 评估MPN恶性概率 520名患者,1739个结节 AUC为0.909(95%CI:0.854~0.946)
Brier评分为0.122
Xin Li等[20] 2019 多发 3D DenseSharp Network 评估AI预测与术后病理结果符合程度 53名患者,108个病灶 恶性肿瘤总体符合率为88.8%,AUC=0.766
Zhanyu Ge等[40] 2005 - CAD 利用三维体积法降低预测肺结节的假阳性率 56名患者 旧特征空间假阳性率:1.61
新特征空间假阳性率:0.37
组合特征空间假阳性率:0.34
Giuseppe Coppini等[41] 2003 - ANNs 检测胸片中的肺结节 65张胸部后前位x线片:45张包含结节,20张无结节 准确率达95.7%
1
Bonney A, Malouf R, Marchal C, et al.Impact of low-dose computed tomography (LDCT) screening on lung cancer-related mortality [J].Cochrane Db Syst Rev, 2022, 8(8): CD013829.
2
中华医学会呼吸病学分会肺癌学组, 中国肺癌防治联盟专家组.肺结节诊治中国专家共识 (2018 年版) [J].中华结核和呼吸杂志,2018, 41(10): 763-771.
3
He YT, Zhang YC, Shi GF, et al.Risk factors for pulmonary nodules in north China: A prospective cohort study [J].Lung Cancer (Amsterdam,Netherlands), 2018, 120: 122-129.
4
钟华, 姚烽, 陈群慧, 等.肺部多发结节的诊断和治疗 [J].中华肿瘤杂志, 2023, 45(6): 455-463.
5
Hattori A, Matsunaga T, Takamochi K, et al.Surgical management of multifocal ground-glass opacities of the lung: correlation of clinicopathologic and radiologic findings [J].Thorac Cardiovasc Surg,2017, 65(2): 142-149.
6
Pan W, Fang X, Zang Z, et al.Diagnostic efficiency of artificial intelligence for pulmonary nodules based on CT scans [J].Am J Transl Res, 2023, 15(5): 3318-3325.
7
Liu P ran, Lu L, Zhang J yao, et al.Application of artificial intelligence in medicine: An overview [J].Current Medical Science, 2021, 41(6):1105-1115.
8
Basu K, Sinha R, Ong A, et al.Artificial intelligence: How is it changing medical sciences and its future? [J].Indian J Dermatol, 2020,65(5): 365-370.
9
郑娟, 徐兴祥.多发性肺结节的临床研究进展 [J].国际呼吸杂志,2021, 41(24).
10
Fu F, Zhang Y, Wen Z, et al.Distinct prognostic factors in patients with stage I non-small cell lung cancer with radiologic part-solid or solid lesions [J].J Thorac Oncol, 2019, 14(12): 2133-2142.
11
Callister ME, Baldwin DR, Akram AR, et al.British Thoracic Society guidelines for the investigation and management of pulmonary nodules[J].Thorax, 2015, 70 Suppl 2: ii1-ii54.
12
Jiang N, Zhang L, Hao Y, et al.Combination of electromagnetic navigation bronchoscopy-guided microwave ablation and thoracoscopic resection: An alternative for treatment of multiple pulmonary nodules [J].Thorac Cancer, 2020, 11(6): 1728-1733.
13
陈丽, 徐培, 唐丽娜, 等.530 例肺结节患者的临床病理特征分析[J].广西医学, 2021, 43(16): 1977-1980.
14
颜世锐, 许萌.以多发肺结节起病的盖尔森基兴奴卡菌感染一例[J/OL].中华临床医师杂志(电子版), 2022, 16(8): 815-817.
15
Taci H, Söyler Y, Uğurman F, et al.An uncommon cause of multiple pulmonary nodules: pulmonary hyalinizing granuloma [J].Hippokratia, 2019, 23(1): 33-36.
16
Ather S, Kadir T, Gleeson F.Artificial intelligence and radiomics in pulmonary nodule management: current status and future applications[J].Clin Radiol, 2020, 75(1):13-19.
17
Marchianò A, Calabrò E, Civelli E, et al.Pulmonary nodules: volume repeatability at multidetector CT lung cancer screening [J].Radiology,2009, 251(3):919-925.
18
Mehta HJ, Ravenel JG, Shaftman SR, et al.The utility of nodule volume in the context of malignancy prediction for small pulmonary nodules [J].Chest, 2014, 145(3):464-472.
19
Yu WS, Hong SR, Lee JG, et al.Three-dimensional ground glass opacity ratio in CT images can predict tumor invasiveness of stage IA lung cancer [J].Yonsei Med J, 2016, 57(5): 1131-1138.
20
Li X, Hu B, Li H, et al.Application of artificial intelligence in the diagnosis of multiple primary lung cancer [J].Thorac Cancer, 2019,10(11): 2168-2174.
21
National Lung Screening Trial Research Team; Aberle DR, Adams AM, Berg CD, et al.Reduced lung-cancer mortality with low-dose computed tomographic screening [J].N Engl J Med, 2011, 365(5):395-409.
22
Armato SG, McLennan G, Bidaut L, et al.The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI):a completed reference database of lung nodules on CT scans [J].Med Phys, 2011, 38(2):915-931.
23
Masood A, Sheng B, Yang P, et al.Automated decision support system for lung cancer detection and classification via enhanced RFCN with multilayer fusion RPN [J].IEEE Trans Ind Informat, 2020, 16(12):7791-7801.
24
Nakaura T, Higaki T, Awai K, et al.A primer for understanding radiology articles about machine learning and deep learning [J].Diagn Interv Imaging, 2020, 101(12): 765-770.
25
Chassagnon G, De Margerie-Mellon C, Vakalopoulou M, et al.Artificial intelligence in lung cancer: current applications and perspectives [J].Jpn J Radiol, 2023, 41(3): 235-244.
26
Peeken JC, Asadpour R, Specht K, et al.MRI-based delta-radiomics predicts pathologic complete response in high-grade soft-tissue sarcoma patients treated with neoadjuvant therapy [J].Radiother Oncol, 2021, 164: 73-82.
27
Ma Y, Li J, Xu X, et al.The CT delta-radiomics based machine learning approach in evaluating multiple primary lung adenocarcinoma[J].BMC Cancer, 2022, 22(1): 949.
28
Chen K, Nie Y, Park S, et al.Development and validation of machine learning-based model for the prediction of malignancy in multiple pulmonary nodules: Analysis from multicentric cohorts [J].Clin Cancer Res, 2021, 27(8): 2255-2265.
29
Zhang L, Shao Y, Chen G, et al.An artificial intelligence-assisted diagnostic system for the prediction of benignity and malignancy of pulmonary nodules and its practical value for patients with different clinical characteristics [J].Front Med (Lausanne), 2023, 10: 1286433.
30
Lv J, Li J, Liu Y, et al.Artificial intelligence-aided diagnosis software to identify highly suspicious pulmonary nodules [J].Front Oncol,2021, 11: 749219.
31
林绍涌, 朱先理, 魏梁锋, 等.ChatGPT 在临床医学实践中的应用进展 [J/OL].中华临床医师杂志 (电子版), 2023, 17(10): 1113-1116.
32
Liu Q, Lv X, Zhou D, et al.Establishment and validation of multiclassification prediction models for pulmonary nodules based on machine learning [J].Clin Respir J, 2024, 18(5): e13769.
33
Takeshita Y, Onozawa S, Katase S, et al.Evaluation of an artificial intelligence U-net algorithm for pulmonary nodule tracking on chest computed tomography images [J].J Int Med Res, 2024, 52(2):03000605241230033.
34
Fan W, Liu H, Zhang Y, et al.Diagnostic value of artificial intelligence based on computed tomography (CT) density in benign and malignant pulmonary nodules: a retrospective investigation [J].Peerj, 2024, 12:e16577.
35
Zhang H, Wang S, Deng Z, et al.Computed tomography-based radiomics machine learning models for prediction of histological invasiveness with sub-centimeter subsolid pulmonary nodules: a retrospective study [J].Peerj, 2023, 11: e14559.
36
Mu J, Kuang K, Ao M, et al.Deep learning predicts malignancy and metastasis of solid pulmonary nodules from CT scans [J].Frontiers in Medicine, 2023, 10: 1145846.
37
Tang TW, Lin WY, Liang JD, et al.Artificial intelligence aided diagnosis of pulmonary nodules segmentation and feature extraction[J].Clin Radiol, 2023, 78(6): 437-443.
38
Zhang K, Wei ZH, Wang X, et al.[The diagnostic value of machinelearning-based model for predicting the malignancy of solid nodules in multiple pulmonary nodules] [J].Zhonghua Wai Ke Za Zhi, 2022,60(6): 573-579.
39
Zhang R, Wei Y, Shi F, et al.The diagnostic and prognostic value of radiomics and deep learning technologies for patients with solid pulmonary nodules in chest CT images [J].BMC Cancer, 2022,22(1): 1118.
40
Ge Z, Sahiner B, Chan HP, et al.Computer-aided detection of lung nodules: false positive reduction using a 3D gradient field method and 3D ellipsoid fitting [J].Med Phys, 2005, 32(8): 2443-2454.
41
Coppini G, Diciotti S, Falchini M, et al.Neural networks for computeraided diagnosis: detection of lung nodules in chest radiograms [J].IEEE Trans Inf Technol Biomed, 2003, 7(4): 344-357.
[1] 李洋, 蔡金玉, 党晓智, 常婉英, 巨艳, 高毅, 宋宏萍. 基于深度学习的乳腺超声应变弹性图像生成模型的应用研究[J]. 中华医学超声杂志(电子版), 2024, 21(06): 563-570.
[2] 杨敬武, 周美君, 陈雨凡, 李素淑, 何燕妮, 崔楠, 刘红梅. 人工智能超声结合品管圈活动对低年资超声医师甲状腺结节风险评估能力的作用[J]. 中华医学超声杂志(电子版), 2024, 21(05): 522-526.
[3] 明昊, 肖迎聪, 巨艳, 宋宏萍. 乳腺癌风险预测模型的研究现状[J]. 中华乳腺病杂志(电子版), 2024, 18(05): 287-291.
[4] 张嘉炜, 王瑞, 张克诚, 易磊, 周增丁. 烧烫伤创面深度智能检测模型P-YOLO的建立及测试效果[J]. 中华损伤与修复杂志(电子版), 2024, 19(05): 379-385.
[5] 叶莉, 杜宇. 深度学习在牙髓根尖周病临床诊疗中的应用[J]. 中华口腔医学研究杂志(电子版), 2024, 18(06): 351-356.
[6] 熊鹰, 林敬莱, 白奇, 郭剑明, 王烁. 肾癌自动化病理诊断:AI离临床还有多远?[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 535-540.
[7] 李伟, 宋子健, 赖衍成, 周睿, 吴涵, 邓龙昕, 陈锐. 人工智能应用于前列腺癌患者预后预测的研究现状及展望[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 541-546.
[8] 黄俊龙, 李文双, 李晓阳, 刘柏隆, 陈逸龙, 丘惠平, 周祥福. 基于盆底彩超的人工智能模型在女性压力性尿失禁分度诊断中的应用[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 597-605.
[9] 莫淇舟, 苏劲, 黄健, 李健维, 李思宁, 柳建军. 智能控压输尿管软镜碎石吸引取石术在直径10~25 mm上尿路结石中的应用[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(05): 497-502.
[10] 苏博兴, 肖博, 李建兴. 2024年美国泌尿外科学会年会结石领域手术治疗相关热点研究及解读[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(04): 303-308.
[11] 莫林键, 杨舒博, 农卫赟, 程继文. 人工智能虚拟数字医师在钬激光前列腺剜除日间手术患教管理中的应用[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(04): 318-322.
[12] 阮星星, 黄智渊, 刘芙香, 狄金明. 从临床医师诊治患者的思路出发撰写临床研究论文[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(04): 397-401.
[13] 李义亮, 苏拉依曼·牙库甫, 麦麦提艾力·麦麦提明, 克力木·阿不都热依木. 机器人与腹腔镜食管裂孔疝修补术联合Nissen 胃底折叠术短期疗效分析[J]. 中华疝和腹壁外科杂志(电子版), 2024, 18(05): 512-517.
[14] 赵毅, 李昶田, 唐文博, 白雪婷, 刘荣. 腹腔镜术中超声主胰管自动识别模型的临床应用[J]. 中华腔镜外科杂志(电子版), 2024, 17(05): 290-294.
[15] 张玮玮, 霍晓川. 人工智能时代医学生批判性思维培养的重要性[J]. 中华脑血管病杂志(电子版), 2024, 18(04): 357-359.
阅读次数
全文


摘要