切换至 "中华医学电子期刊资源库"
高级检索
中华肺部疾病杂志(电子版)

中华肺部疾病杂志(电子版) ›› 2024, Vol. 17 ›› Issue (02) : 242 -246. doi: 10.3877/cma.j.issn.1674-6902.2024.02.014

论著

采用AI的CT影像特征对肺结节良恶性鉴别的价值分析
解良婕1, 王剑2,†(), 阳韬2,†()   
  1. 1. 212013　镇江，江苏大学医学院
    2. 212002　镇江，江苏大学附属人民医院（镇江市第一人民医院）呼吸及危重症学科
  • 收稿日期:2023-11-13 出版日期:2024-04-25
  • 通信作者: 王剑, 阳韬
  • 基金资助:
    镇江市社会发展重大项目(SH2020047); 镇江市社会发展指导项目(FZ2019016)

Value analysis of CT image features based on AI in differential diagnosis of benign and malignant Pulmonary nodules

Liangjie Xie1, Jian Wang2,†(), Tao Yang2,†()   

  1. 1. School of Medicine, Jiangsu University, Zhenjiang 212013, China
    2. Department of Pulmonary and Critial Care Medicine, Affiliated People′s Hospital of Jiangsu University, Zhenjiang 212002, China
  • Received:2023-11-13 Published:2024-04-25
  • Corresponding author: Jian Wang, Tao Yang
全文 ( ) 下载PDF ( ) 导出引用
引用本文：

解良婕, 王剑, 阳韬. 采用AI的CT影像特征对肺结节良恶性鉴别的价值分析[J]. 中华肺部疾病杂志(电子版), 2024, 17(02): 242-246.

Liangjie Xie, Jian Wang, Tao Yang. Value analysis of CT image features based on AI in differential diagnosis of benign and malignant Pulmonary nodules[J]. Chinese Journal of Lung Diseases(Electronic Edition), 2024, 17(02): 242-246.

目的

采用人工智能(artificial intelligence, AI)的胸部低剂量计算机断层扫描(low-dose computed tomography, LDCT)影像特征分析肺结节良恶性的危险因素，建立肺结节良恶性预测模型。

方法

选择2021年1月至2022年12月我院收治的肺结节患者240例，收集临床资料，采用LDCT及AI鉴别肺结节，进行多因素分析，筛选肺结节良恶性的危险因素，建立二元Logistics回归模型，比较AI、影像医师及预测模型的诊断价值。

结果

最小CT值、直径R、毛刺征、血管穿行征、纯磨玻璃结节、部分实性结节是影响肺结节良恶性的危险因素(P<0.05)。二元Logistics回归模型为logit(P)＝－2.905＋(0.93×直径R)＋(1.572×血管穿行)＋(1.346×毛刺征)＋(1.755×纯磨玻璃结节)＋(2.25×部分实性结节)－(0.001×最小CT值)，AI、影像医师及预测模型鉴别的灵敏度分别为89.86%、81.88%、73.19%，特异度分别为32.35%、55.88%、73.53%，阳性似然比分别为1.328、1.856、2.762，阴性似然比分别为0.314、0.324、0.365，曲线下面积(area under the curve, AUC)分别为0.611、0.689、0.789。

结论

联合肺结节形态特征及基于AI的CT定量参数回归模型对肺结节良恶性的诊断价值优于AI及影像医师，具有临床意义。

关键词: 肺结节, 危险因素, 人工智能, 预测模型
Objective

To analyze the risk factors of benign and malignant lung nodules and establish a predictive model by the low dose computed tomography (LDCT) images and artificial intelligence (AI).

Methods

The clinical data of 240 patients, who were admitted to our hospital from January 2021 to December 2022 and had definitive pathological results, were retrospectively analyzed. Univariate and multivariate analyses were performed to identify the risk factors associated with benign and malignant pulmonary nodules. A binary logistic regression model was established, and the diagnostic efficiency of AI, radiologists, and prediction models was evaluated.

Results

Risk factors for distinguishing benign and malignant pulmonary nodules included minimum CT value, diameter R, spiculation sign, vascular penetration sign, pure ground glass nodules, and partially solid nodules(P<0.05). The regression model constructed was logit(P)=-2.905+ (0.93×diameter R)+ (1.572×vascular penetration)+ (1.346×spiculation sign)+ (1.755×pure ground glass nodules)+ (2.25×partially solid nodules)-(0.001×minimum CT value). The sensitivity of AI, radiologists, and the prediction model was 89.86%, 81.88%, and 73.19% respectively, with corresponding specificities of 32.35%, 55.88%, and 73.53%. The positive likelihood ratios was 1.328, 1.856, and 2.762, and the negative likelihood ratios are 0.314, 0.324, and 0.365. The AUC values was 0.611, 0.689, and 0.789 respectively.

Conclusion

The diagnostic performance of the regression model, which combines pulmonary nodule morphological features and AI-based CT quantitative parameters, in distinguishing between benign and malignant pulmonary nodules is moderate, but superior to that of AI alone or individual radiologists.

Key words: Pulmonary nodules, Risk factors, Artificial intelligence, Predictive model
表1 肺结节良恶性单因素分析[(±s)，n(%)]
临床资料 良性结节(n＝102) 恶性结节(n＝138) P值
性别     0.001
男 58(56.86) 50(36.23)  
女 44(43.14) 88(63.77)  
年龄(岁) 57.65±10.58 59.17±11.16 0.265
吸烟 25(24.51) 23(16.67) 0.133
结节体积(mm3) 1 680.95±3 023.76 1 564.61±2 154.82 0.156
平均CT值(Hu) －275.22±316.36 －456.59±238.18 0.000
最小CT值(Hu) －744.72±344.28 －910.99±688.06 0.004
最大CT值(Hu) 186.00±532.26 49.38±396.29 0.030
毛刺征 12(11.76) 30(21.74) 0.044
血管穿行 4(3.92) 27(19.57) 0.000
胸膜牵拉 9(8.82) 19(13.77) 0.238
分叶征 3(2.94) 12(8.70) 0.121
纯磨玻璃结节 33(32.35) 65(47.10) 0.022
部分实性结节 9(8.82) 43(31.16) 0.000
实性结节 60(58.83) 30(21.74) 0.000
平均直径(mm) 11.73±6.44 12.15±5.64 0.290
直径R(mm)     0.036
8<R≤30 70(68.63) 111(80.43)  
R≤8 32(31.37) 27(19.57)  
表2 肺结节良恶性Logistic多因素分析
影响因素 B值 标准误 瓦尔德 OR值 HR(95%CI) P值
直径R(mm) 0.930 0.360 6.684 2.534 (1.252，5.126) 0.010
血管穿行 1.572 0.646 5.923 4.818 (1.358，17.090) 0.015
毛刺征 1.346 0.453 8.823 3.843 (1.581，9.343) 0.003
纯磨玻璃结节 1.755 0.377 21.652 5.782 (2.761，12.108) 0.000
部分实性结节 2.250 0.487 21.372 9.490 (3.66，24.64) 0.000
最小CT值(Hu) －0.001 0.001 4.632 0.999 (0.998，1.000) 0.031
常量 －2.905 0.621 21.903 0.055 / /
表3 AI、影像学及预测模型的肺结节诊断比较(%)
诊断方式 AUC 敏感性 特异性 阳性似然值 阴性似然值
AI 0.611 89.86 32.35 1.328 0.314
影像医师 0.689 81.88 55.88 1.856 0.324
预测模型 0.789 73.19 73.53 2.762 0.365
1
高嘉营，金发光. 肺癌自身抗体在肺癌诊断中的研究进展[J/CD]. 中华肺部疾病杂志(电子版), 2023, 16(5): 739-741.
2
Oudkerk M, Liu S, Heuvelmans MA, et al. Lung cancer LDCT screening and mortality reduction-evidence, pitfalls and future perspectives[J]. Nat Rev Clin Oncol, 2021, 18(3): 135-151.
3
Bach PB, Mirkin JN, Oliver TK, et al. Benefits and harms of CT screening for lung cancer: a systematic review[J]. JAMA, 2012, 307(22): 2418-2429.
4
Jonas DE, Reuland DS, Reddy SM, et al. Screening for lung cancer with low-dose computed tomography: Updated evidence report and systematic review for the US preventive services task force[J]. JAMA, 2021, 325(10): 971-987.
5
Mcwilliams A, Tammemagi MC, Mayo JR, et al. Probability of cancer in pulmonary nodules detected on first screening CT[J]. N Engl J Med, 2013, 369(10): 910-919.
6
杨　丽，钱桂生. 肺结节临床精准诊断的新理念[J/CD]. 中华肺部疾病杂志(电子版), 2022, 15(1): 1-5.
7
Hawkins S, Wang H, Liu Y, et al. Predicting malignant nodules from screening CT scans[J]. J Thorac Oncol, 2016, 11(12): 2120-2128.
8
Zhao W, Yang J, Sun Y, et al. 3D deep learning from CT scans predicts tumor invasiveness of subcentimeter pulmonary adenocarcinomas[J]. Cancer Res, 2018, 78(24): 6881-6889.
9
Ardila D, Kiraly AP, Bharadwaj S, et al. End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography[J]. Nature Medicine, 2019, 25(6): 954-961.
10
Venkadesh KV, Setio AaA, Schreuder A, et al. Deep learning for malignancy risk estimation of pulmonary nodules detected at low-dose screening CT[J]. Radiology, 2021, 300(2): 438-447.
11
吴阶平医学基金会模拟医学部胸外科专委会. 人工智能在肺结节诊治中的应用专家共识[J]. 中国肺癌杂志，2022, 25(4): 219-225.
12
Pei Q, Luo Y, Chen Y, et al. Artificial intelligence in clinical applications for lung cancer: diagnosis, treatment and prognosis[J]. Clin Chem Lab Med, 2022, 60(12): 1974-1983.
13
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.
14
Li C, Lei S, Ding L, et al. Global burden and trends of lung cancer incidence and mortality[J]. Chin Med J (Engl), 2023, 136(13): 1583-1590.
15
Seijo L M, Peled N, Ajona D, et al. Biomarkers in lung cancer screening: Achievements, promises, and challenges[J]. J Thor Oncol, 2019, 14(3): 343-357.
16
Chabon JJ, Hamilton EG, Kurtz DM, et al. Integrating genomic features for non-invasive early lung cancer detection[J]. Nature, 2020, 580(7802): 245-251.
17
Chamberlin J, Kocher MR, Waltz J, et al. Automated detection of lung nodules and coronary artery calcium using artificial intelligence on low-dose CT scans for lung cancer screening: accuracy and prognostic value[J]. BMC Med, 2021, 19(1): 55.
18
Wang X, Gao M, Xie J, et al. Development, validation, and comparison of image-based, clinical feature-based and fusion artificial intelligence diagnostic models in differentiating benign and malignant pulmonary ground-glass nodules[J]. Front Oncol, 2022, 12: 892890.
19
Yoshida K, Gowers KHC, Lee-Six H, et al. Tobacco smoking and somatic mutations in human bronchial epithelium[J]. Nature, 2020, 578(7794): 266-272.
20
Daniel M, Keefe FJ, Lyna P, et al. Persistent smoking after a diagnosis of lung cancer is associated with higher reported pain levels[J]. J Pain, 2009, 10(3): 323-328.
21
Chen WQ, Zhang SW, Zou XN, et al. Cancer incidence and mortality in china, 2006[J]. Chin J Cancer Res, 2011, 23(1): 3-9.
22
Siegel RL, Miller KD, Fuchs HE, et al. Cancer statistics, 2022[J]. CA Cancer J Clin, 2022, 72(1): 7-33.
23
Liang ZR, Ye M, Lv FJ, et al. Differential diagnosis of benign and malignant patchy ground-glass opacity by thin-section computed tomography[J]. BMC Cancer, 2022, 22(1): 1206.
24
Ragavan M, Patel MI. The evolving landscape of sex-based differences in lung cancer: a distinct disease in women[J]. Eur Respir Rev, 2022, 31(163). 210100.
25
Macrosty CR, Rivera MP. Lung cancer in women: A modern epidemic[J]. Clin Chest Med, 2020, 41(1): 53-65.
26
Li X, Zhang W, Yu Y, et al. CT features and quantitative analysis of subsolid nodule lung adenocarcinoma for pathological classification prediction[J]. BMC Cancer, 2020, 20(1): 60.
27
Chu ZG, Li WJ, Fu BJ, et al. CT Characteristics for predicting invasiveness in pulmonary pure ground-glass nodules[J]. Am J Roentgenol, 2020, 215(2): 351-358.
28
Jacob M, Romano J, Ara Jo D, et al. Predicting lung nodules malignancy[J]. Pulmonol, 2022, 28(6): 454-460.
29
Gao Y, Chen Y, Jiang Y, et al. Artificial intelligence algorithm-based feature extraction of computed tomography images and analysis of benign and malignant pulmonary nodules[J]. Computat Intell Neurosci, 2022, 2022: 5762623.
30
白　静. 球形肺炎的CT诊断价值与鉴别诊断分析[J]. 中国医疗器械信息，2019, 25(20): 1-2＋188.
31
唐　志，周建国. 非实性肺结节CT影像收缩力表现与侵袭性对照分析[J]. 医学影像学杂志，2023, 33(3): 431-434.
[1] 伯小皖, 郭乐杭, 余松远, 李明宙, 孙丽萍. 甲状腺结节人工智能自动分割和分类系统的建立和验证[J]. 中华医学超声杂志（电子版）, 2024, 21(03): 304-309.
[2] 刘韩, 王胰, 舒庆兰, 彭博, 尹立雪, 谢盛华. 基于深度学习的超声心动图三尖瓣反流严重程度智能评估方法研究[J]. 中华医学超声杂志（电子版）, 2024, 21(02): 121-127.
[3] 成汉林, 史中青, 戚占如, 王小贤, 曾子炀, 单淳劼, 钱隼南, 罗守华, 姚静. 基于深度学习的超声心动图动态图像切面识别研究[J]. 中华医学超声杂志（电子版）, 2024, 21(02): 128-136.
[4] 赫兰, 杨泽堃, 张颖, 王玉东, 陈伟导, 王一同, 申锷. 双输入BCNN-ResNet模型对超声颈动脉斑块稳定性的分类诊断价值[J]. 中华医学超声杂志（电子版）, 2024, 21(02): 137-142.
[5] 丁关棣, 黄云, 曹震, 刘刚. 胃癌根治术后感染性并发症预测：基于真实世界数据的Nomogram模型开发与验证[J]. 中华普外科手术学杂志(电子版), 2024, 18(03): 261-266.
[6] 张阳, 纽燕娜, 常丽蓉, 唐国华, 赵萍. ERAS理念下肝棘球蚴病术后并发症风险预测模型构建[J]. 中华普外科手术学杂志(电子版), 2024, 18(03): 287-290.
[7] 高佳, 姜吉询, 王鑫, 吴婷, 向江明. 病理性乳头溢液单中心临床分析附168例报告[J]. 中华普外科手术学杂志(电子版), 2024, 18(03): 323-326.
[8] 黄海, 程必盛, 黄健. 2024年欧洲泌尿外科学会年会：前列腺癌研究的前沿探索与未来趋势[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(03): 202-207.
[9] 杨龙雨禾, 王跃强, 招云亮, 金溪, 卫娜, 杨智明, 张贵福. 人工智能辅助临床决策在泌尿系肿瘤的应用进展[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(02): 178-182.
[10] 钟知足, 陶波圆, 曾纪晓, 吴强, 王哲, 余家康, 刘斐, 徐晓钢, 兰梦龙, 梁子建, 李燕秋, 刘凤君. 儿童腹腔镜胆总管囊肿根治术后并发症危险因素分析[J]. 中华腔镜外科杂志(电子版), 2024, 17(02): 83-89.
[11] 杨竞, 周光文. 肝硬化门静脉高压症治疗后再出血危险因素分析及预测模型构建[J]. 中华肝脏外科手术学电子杂志, 2024, 13(03): 296-301.
[12] 彭子洋, 王志博, 王丹, 彭浩茜, 王蕾, 彭薇, 王娟娟, 李宇, 刘学民, 吴荣谦, 向俊西, 吕毅. 智能化辅助图像实时去雾技术在腹腔镜胆囊切除术中的应用[J]. 中华肝脏外科手术学电子杂志, 2024, 13(03): 328-333.
[13] 赵倩, 刘文超, 李玺琳, 章邱东. 老年急性脑梗死诱发胃肠损伤的风险因素分析及模型构建[J]. 中华消化病与影像杂志(电子版), 2024, 14(03): 213-217.
[14] 邓仙裕, 罗钰璇, 张溱乐, 余展鹏, 彭亮. 自发性脑出血重症患者30 d死亡风险预测模型的建立及验证[J]. 中华脑血管病杂志(电子版), 2024, 18(02): 121-128.
[15] 赖敏, 林碧玲, 王亚文, 林玉禧, 欧开雅. 中青年脑卒中患者重返工作的影响因素分析及预测模型构建[J]. 中华脑血管病杂志(电子版), 2024, 18(02): 145-152.
阅读次数
全文


摘要

版权所有 中华医学会 中华医学电子音像出版社有限责任公司  京ICP备14006079号-1  网络出版服务许可证：(署)网出证(京)字第075号